+ a . rn ee ; PIE A > 7 = u - +“ led . + be} ‘ ’ - ” LA ~ m. = BE LE 7 dm. i : ; . = ee : "= = © AR | .. - , a _ , -— - : ~ ‘ .-< : ‘ .- Pr - … . D ' i y A vi at = gl! fo «ht mt 4 i ar TL — er Eure wir as 5 FOR THE PEOPEE FOR EDVCATION FOR SCIENCE LIBRARY OF THE AMERICAN MUSEUM OF NATURAL HISTORY 5 if NL D vie U it ue Run un i. if Woy. vun i rah | Rare es = POS 5 In L 5 * a RES, Der Oey J N | 7 2 q'4 à JAN JE ro æ yi va ae AAC À NS Lit i (ur La igs | ini Te Lee sr HE 7 9 ig ul br | Ly py rn en Bei IN ar ag”! ie ay aes Wi 4% i ir 4 wil bs da "il i f à x i Y N i i ry Fi à } |: 2 = i f \ i å = 4 , LU ” ' | ] | | 4 L | dl a RE vf i ‘ i i 1 A t AMET 4" i i" La ie > Pa tA) ur ‘ x L på E SNE F l x an am i ivy % "TON i am = auf, N LJ ’ A i i werd bi | td fer li | A 7 2} f i ¥ Å i : di RU Sr EN | rh AE Kata Pav NE rå EN Ta En | AE OUT ha ete ie 4 HR + 5 i vi À à à “ry LA DET KONGELIGE DANSKE VIDENSKABERNES SELSKABS SKRIFTER OTTENDE REKKE NATURVIDENSKABELIG 0G MATHEMATISK AFDELING FEMTE BIND MED 57 TAVLER KØBENHAVN HOVEDKOMMISSIONÆR: ANDR. FRED. HØST & SØN, KGL. HOF-BOGHANDEL BIANCO LUNOS BOGTRYKKERI 1918—24 YN bel i er DH ti ' NUS CUNT RO TRS AE YROTETH TERIA ,, | tn NN à 25-10,0659- OS KITE INDHOLD Side Fortesnelse over Selskabets" Medlemmer Juni 1924 "%.............................. V—XVII 1. Bjerrum, Niels und Kirschner, Aage: Die Rhodanide des Goldes und das freie Rhodan. MimEiMetinanvanecuber das Goldehlond a... se cre oe 1— 77 One -Jensens Se The lactic acid Bacteria. With 51/Plates -. .-.................... 79—197 3. Brünnich Nielsen, K.: Zoantharia from Senone and Paleocene Deposits in Denmark LEE Sierre, MONNIER ES SRE ER Jon EU PE PRET . 199—233 4. Petersen, Axel: Bidrag til de danske Simuliers Naturhistorie. Med 2 Tavler, 53 PLD DIL RO TOR PR ne Ses RE 235—341 FORTEGNELSE OVER DET KONGELIGE DANSKE VIDENSKABERNES: SELSKABS: MEDLEMMER Juni 1924 Protektor: Hans Majestet Kongen. es President: VILH. THOMSEN. Formand for den hist.-filos. Klasse: Fr. BUHL. Formand for den naturv.-math. Klasse: S. P. L. SØRENSEN. Sekretær: MARTIN KNUDSEN. Redaktor: DINES ANDERSEN. Kasserer: W. L. JOHANNSEN. Kassekommissionen. H. HOFFDING. Kr. ERSLEV. Jee Weve ENSENe J. T. HJELMSLEV. Revisorer. Tu. J. M. MADSEN. C. HANSEN OSTENFELD. Kommissionen til Registrering af litterære Kilder til dansk Historie. JOH. STEENSTRUP. Kr. ERSLEV. H. O. LANGE. Kommissionen til Undersøgelse af de i dansk Privateje bevarede Kilder til dansk Historie. JOH. STEENSTRUP. Kr. ERSLEV. HO ANGE AAGE FRIIS. Udvalget vedrorende Union Académique Internationale. J: LL. HEWERG. OTTO JESPERSEN. Cur. BLINKENBERG. Udvalg for den internationale Katalog over naturvidenskabelige Arbejder. L. KOLDERUP ROSENVINGE. V. HENRIQUES. S. P. L. SORENSEN. MARTIN KNUDSEN. J. T, HJELMSLEV: TH. MORTENSEN. Udvalg for Deltagelse i internationale vulkanologiske Undersøgelser. K. Pryrz. S. P. L. SØRENSEN. M. KNUDSEN. O. B. BoGGILp. INDENLANDSKE MEDLEMMER THOMSEN, VILHELM LupviG PETER, Dr. phil., fh. Professor i sammenlignende Sprog- videnskab ved Københavns Universitet, Ridder af Elefanten, Storkors af Danne- brog og Dannebrogsmand, dekoreret med Fortjenstmedaillen i Guld, med det finske Frihedskors 1. Kl. og Storkorset af Finlands hvide Rose, med den preus- siske Orden Pour le Mérite, og den preussiske Rode Orns Orden 2. Klasse med Stjerne, Selskabets Præsident. TorsøE, HALDOR FREDERIK AXEL, Dr. phil., fh. Direktør for Arbejds- og Fabriktil- synet, Kommandør?” af Dannebrog og Dannebrogsmand, dekoreret med Fortjenst- medaillen i Guld. STEENSTRUP, JOHANNES CHRISTOPHER HAGEMANN REINHARDT, Dr. jur. & phil., fh. Pro- fessor Rostgardianus i nordisk Historie og Antikviteter ved Københavns Uni- versitet, Kommandør?” af Dannebrog og Dannebrogsmand, Kommandør af Nord- stjernen, Ridder af Æreslegionen. GERTZ, MARTIN CLARENTIUS, Dr. phil., fh. Professor i klassisk Filologi ved Køben- havns Universitet, Kommandør” af Dannebrog og Dannebrogsmand, Kommandør af den italienske Kroneorden og af Nordstjernen. HEIBERG, JOHAN LunviG, Dr. phil., litt., med. & sc., Professor i klassisk Filologi ved Københavns Universitet, Kommandør” af Dannebrog og Dannebrogsmand, Kommandør af St. Olavs Ordenen og af den italienske Kroneorden, Officer af Æreslegionen. HørrninG, HARALD, Dr. phil., jur., sc. & litt., fh. Professor i Filosofi ved Københavns Universitet, Storkors af Dannebrog og Dannebrogsmand, Kommandør af St. Olavs Ordenen, af Nordstjernen og af Æreslegionen, Officier de l’instruction publique, dekoreret med det finske Frihedskors 1. Kl., det danske Røde Kors” Hæderstegn, med frivilligt Sygeplejevæsens Fortjenstmedaille, det østrigske Røde Kors’ Stjerne, det preussiske Røde Kors” Bryststjerne og Broncemedaille, det russiske Røde Kors' Udmærkelsestegn og med den tyrkiske Røde Halvmaanes Medaille. KROMAN, KRISTIAN FREDERIK VILHELM, Dr. phil., fh. Professor i Filosofi ved Køben- havns Universitet, Ridder af Dannebrog og Dannebrogsmand. MULLER, PETER Erasmus, Dr. phil., Kammerherre, Hofjægermester, fh. Overførster for anden Inspektion, Storkors af Dannebrog og Dannebrogsmand, dekoreret med Erindringstegnet om Kong Christian IX's og Dronning Louises Guldbryllup Ville og Frederik VIII’s Mindetegn, Kommandør af St. Olavs Ordenen, af den russiske St. Annaorden, af den spanske Carl III’s Orden med Stjerne og af den græske Frelserens Orden, dekoreret med den preussiske Røde Orns Orden 2. KI. ErsLev, Kristian Sorus Auaust, Dr. phil., fh. Rigsarkivar, Kommandør" af Dannebrog og Dannebrogsmand, Kommander af St. Olavsordenen. Boas, JoHAN Erik Vesti, Dr. phil., Professor i Zoologi ved den kgl. Veterinær- og Landbohgjskole, Ridder af Dannebrog og Dannebrogsmand, Kommander af St. Olavs-Ordenen. PETERSEN, OTTO GEORG, Dr. phil., fh. Professor i Botanik ved den kgl. Veterinær- og Landbohøjskole, Ridder af Dannebrog og Dannebrogsmand. Prytz, PETER KRISTIAN, fh. Professor i Fysik ved den Polytekniske Læreanstalt, Kommandør” af Dannebrog og Dannebrogsmand. SALOMONSEN, CARL JULIUS, Dr. med. & sc., fh. Professor i Pathologi ved Københavns Universitet, Kommandør” af Dannebrog og Dannebrogsmand, Kommandør af den russiske St. Stanislausorden, af den svenske Vasaorden og af Æreslegionen, Ridder af Nordstjernen og af St. Olavs Ordenen, Officier de l'instruction pu- blique, dekoreret med den preussiske Kroneorden 2. Kl. JONSSON, Finnur, Dr. phil. & litt. isl., Professor i nordisk Filologi ved Københavns Universitet, Ridder af Dannebrog og Dannebrogsmand, Storridder af den islandske Falk, Kommandør af Nordstjernen, Ridder af St. Olavs Ordenen. MÜLLER, SorHus OTTO, Dr. phil., fh. Direktør for Nationalmuseets første Afdeling, Kommandør” af Dannebrog og Dannebrogsmand, Kommandør af St. Olavs Ordenen, af Nordstjernen og af den italienske St. Mauritius og Lazarusorden, Ridder af Æreslegionen. BERGH, RuDoLPH Sopuus, Dr. phil., fh. Docent i Histologi ved Københavns Universitet. JOHANNSEN, WILHELM LupviG, Dr. med., phil. & bot. et zool., Professor i Plante- fysiologi ved Københavns Universitet, Kommandør” af Dannebrog og Danne- brogsmand, Kommandor af Nordstjernen og af den franske Orden Mérite agricole, dekoreret med det preussiske Rode Kors’ Bryststjerne og Broncemedaille og med det ostrigske Rode Kors’ Ærestegn af 1. Klasse, Selskabets Kasserer. JESPERSEN, JENS OTTO Harry, Dr. phil. & litt., Professor i engelsk Sprog og Litte- ratur ved Kobenhavns Universitet, Ridder af Dannebrog. Nyrop, KRISTOFFER, Dr. phil., Professor i romansk Sprog og Litteratur ved Koben- havns Universitet, Kommandør” af Dannebrog og Dannebrogsmand, Kommandør af den belgiske Kroneorden, af Æreslegionen, af den italienske Kroneorden og af den spanske Isabella den Katholskes Orden, Officier de l'instruction publique, dekoreret med den rumænske Fortjenstmedaille i Guld. BANG, BERNHARD Laurirs FREDERIK, Dr. med., fh. Veterinærfysikus, fh. Professor i Veterinær-Lægevidenskab ved den-kgl. Veterinær- og Landbohøjskole, Storkors af Dannebrog og Dannebrogsmand, Kommandør af Nordstjernen, af St. Olavs =k Ordenen og af Finlands hvide Rose, Komtur af 1. Kl. med Stjerne af den sach- siske Albrechtsorden. JUEL, CHRISTIAN Sopuus, Dr. phil., Professor i Mathematik ved den Polytekniske Lære- anstalt i Kobenhayn, Ridder af Dannebrog og Dannebrogsmand. Bunt, FRANTZ PETER WiLLiAM, Dr. phil. & theol., fh. Professor i semitisk-orientalsk Filologi ved Københavns Universitet, Kommandør" af Dannebrog og Dannebrogs- mand, Kommandør af den græske Frelserorden, Ridder af Nordstjernen og af Kongeriget Sachsens Civil Fortjeneste Orden, Officier de l'instruction publique, Formand i Selskabets historisk-filosofiske Klasse. ROSENVINGE, JANUS LAURITS ANDREAS KOLDERUP, Dr. phil., Professor i Botanik ved Kobenhavns Universitet, Ridder af Dannebrog. DREYER, JOHAN LupviG Emit, Dr. phil. & sc., fh. Director of the Armagh Observatory, Irland, Ridder af Dannebrog. _ RAUNKIÆR, CHRISTEN, fh. Professor i Botanik ved Københavns Universitet. DRACHMANN, ANDERS Born, Dr. phil., Professor i klassisk Filologi ved Københavns Universitet, Ridder af Dannebrog og Dannebrogsmand. Hupe, Kristian Karz TuLrinius, Dr. phil., Rektor ved Metropolitanskolen, Lektor ved Kobenhavns Universitet, Ridder af Dannebrog. HENRIQUES, VALDEMAR, Dr. med., Professor i Fysiologi ved Kobenhavns Universitet. JENSEN, CARL OLur, Dr. med. & med. vet., Professor, Forstander for den kgl. Vete- rinær- og Landbohojskoles Serumlaboratorium, Veterinærfysikus, Kommandør?” af Dannebrog og Dannebrogsmand, Storridder af den islandske Falk, Kommandor af St. Olavs Ordenen og af Vasaordenen, Ridder af Nordstjerneordenen, Officer af den sachsiske Albrechtsorden, dekoreret med Ordenen »Polonia Restituta« 3. Klasse. PEDERSEN, HOLGER, Dr. phil., Professor i sammenlignende Sprogvidenskab ved Koben- havns Universitet, Ridder af Dannebrog. LANGE, Hans OSTENFELDT, Dr. phil., fh. Overbibliotekar ved det kongelige Bibliotek i Kobenhavn, Lektor i Ægyptologi ved Kobenhavns Universitet, Ridder af Danne- brog og Dannebrogsmand, Komtur af den meklenborgske Griforden, Ridder af St. Olavs Ordenen. SORENSEN, SOREN PETER LAURITZ, Dr. phil. & med., Professor, Direktor for Carlsberg- laboratoriets kemiske Afdeling, Kobenhayn, Ridder af Dannebrog og Dannebrogs- mand, Ridder af Æreslegionen, Formand for Selskabets naturvidenskabelig- mathematiske Klasse. JENSEN, JOHAN LupviG WILLIAM VALDEMAR, Dr. phil., fh. Overingenior, Ridder af Dannebrog og Dannebrogsmand. ANDERSEN, DINES, Dr. phil., Professor i indisk-osterlandsk Filologi ved Københavns Universitet, Selskabets Redaktor. KNUDSEN, MARTIN Hans CHRISTIAN, Dr. phil., Professor i Fysik ved Københavns Universitet og den Polytekniske Lereanstalt, Ridder af Dannebrog og Danne- Il X brogsmand, Kommandør af Finlands hvide Rose, Ridder af St. Olavs Ordenen, Selskabets Sekretær. MADSEN, THORVALD JOHANNES Marius, Dr. med., Direkter for Statens Seruminstitut, Kommandør” af Dannebrog og Dannebrogsmand, Storofficer af den italienske Kroneorden, Kommander af Nordstjernen, af Æreslegionen og af Ordenen “The British Empire”, Officer af den belgiske Kroneorden, Ridder af St. Olavs Ordenen og af den portugisiske San Thiago Orden, dekoreret med den preus- siske Rode Orns Orden 4. KI, med det danske Rode Kors’ Hæderstegn, med det preussiske Rode Kors’ Bryststjerne og Broncemedaille, med det russiske Rode Kors’ Udmærkelsestegn, med det østrigske Rode Kors’ Officersærestegn og Hæderstegn, med Ordenen »Polonia Restituta« 2. Kl., med fransk Médaille de Reconnaissance 2. KI. og med Rode Halvmaanes Medaille. BLINKENBERG, CHRISTIAN SORENSEN, Dr. phil., Professor i Arkæologi ved Københavns Universitet, Ridder af Dannebrog, Kommander af den spanske Isabella den Katholskes Orden. VEDEL, VALDEMAR, Dr. phil., Professor i almindelig Litteraturvidenskab ved Koben- havns Universitet, Ridder af St. Olavs Ordenen. SANDFELD, Kristian, Dr. phil., Professor i romanske Sprog ved Københavns Uni- versitet. SARAUW, CHRISTIAN PREBEN EMIL, Dr. phil., Professor i tysk Sprog og Litteratur ved Kobenhavns Universitet. Bock, JOHANNES CARL, Dr. med., Professor i Farmakologi ved Københavns Univer- sitet, Kommandør” af Dannebrog og Dannebrogsmand. BRONSTED, JOHANNES NICOLAUS, Dr. phil., Professor i Kemi ved Københavns Universitet. HJELMSLEV, JOHANNES TROLLE, Dr. phil., Professor i Mathematik ved Kobenhavns Universitet. NIELSEN, NIELS, Dr. phil., Professor i Mathematik ved Kobenhavns Universitet, Ridder af Dannebrog. . PETERSEN, CARL GEORG JOHANNES, Dr. phil., & jur. & sc., Direktor for Dansk biologisk Station, Ridder af Dannebrog og Dannebrogsmand, Kommandør af den russiske St. Stanislausorden, Ridder af St. Olavs Ordenen og af Nordstjernen. POULSEN, VALDEMAR, Dr. phil., Ingeniør, Kommandør?” af Dannebrog, dekoreret med Fortjenstmedaillen i Guld med Krone. BJERRUM, NIELS JANNIKSEN, Dr. phil., Professor i Kemi ved den kgl. Veterinær- og Landbohojskole, Ridder af Dannebrog, Ridder af St. Olavs Ordenen. FIBIGER, JOHANNES ANDREAS Gris, Dr. med., Professor i pathologisk Anatomi ved Kobenhayns Universitet, Ridder af Dannebrog og Dannebrogsmand. KROGH, SCHACK AUGUST STEENBERG, Dr. phil. & jur., Professor i Dyrefysiologi ved Kobenhavns Universitet. NoRLUND, Niers Erik, Dr. phil., Professor i Mathematik ved Københavns Uni- versitel, Direktor for Gradmaalingen, Ridder af Nordstjernen. XI OSTENFELD, Cart Emit Hansen, Dr. phil., Professor i Botanik ved Københavns Universitet, Ridder af Dannebrog, Kommandor af Finlands hvide Rose, dekoreret med den tunesiske Nichan-el-Iftikhar Orden 2. Kl. Bour, Niets Davin Henrik, Dr. phil. & sc., Professor i theoretisk Fysik-ved Koben- havns Universitet, Ridder af Dannebrog, Ridder af St. Olavs Ordenen. GRÖNBECH, VILH. PETER, Dr. phil. & theol., Professor extraord. i Religionshistorie ved Kobenhavns Universitet. PEDERSEN, PEDER ÖLUF, Professor i Elektroteknik ved og Direktor for Polyteknisk Lereanstalt, Ridder af Dannebrog og Dannebrogsmand, dekoreret med Fortjenst- medaillen i Guld, Kommander af St. Olavs Ordenen. CHRISTENSEN, ARTHUR EMANUEL, Dr. phil., Professor extraord. i iransk Filologi ved Københavns Universitet. Bour, HARALD AUGUST, Dr. phil., Professor i Mathematik ved Polyteknisk Læreanstalt. SCHMIDT, Ernst JOHANNES, Dr. phil., Direktor for Carlsberglaboratoriets fysiologiske Afdeling, Ridder af Dannebrog og Dannebrogsmand, Ridder af St. Olavs Ordenen, Officer af Æreslegionen og af den siamesiske Krone Orden, dekoreret med den tunesiske Nichan-el-Iftikhar Orden 3. Kl. og med den italienske Jordskelvs- medaille. WESENBERG-LUND, CARL JORGENSEN, Dr. phil., Professor extraord. i Ferskvandsbiologi ved Kobenhavns Universitetet, Leder af Universitetets biologiske Ferskvands- Laboratorier. BoGGILD, Ove BALTHASAR, Professor i Mineralogi ved Københavns Universitet. HERTZSPRUNG, EJNAR, Dr. math. et phys., Professor, Underdirektor ved Observatoriet i Leiden. Frus, AAGE, Dr. phil., Professor Rostgardianus i nordisk Historie ved Kobenhavns Universitet. POULSEN, FREDERIK, Dr. phil., Inspektor ved Ny Carlsberg Glyptothek, Ridder at Nordstjernen og af Æreslegionen, Officier de l'instruction publique, dekoreret med den tunesiske Nichan-el-Iftikhar Orden 3. Kl. ELLERMANN, VILHELM, Dr. med., Professor i Retsmedicin ved Kobenhavns Universitet. MORTENSEN, OLE THEODOR JENSEN, Dr. phil., Inspektor ved Københavns Univer- sitets Zoologiske Museum, Officer af den siamesiske Kroneorden. ANDERSEN, VILHELM Rasmus ANDREAS, Dr. phil., Professor i nordisk Litteratur ved Kobenhavns Universitet, Ridder af Dannebrog og Dannebrogsmand, Kommandor af St. Olavs Ordenen og af Nordstjernen. THALBITZER, CARL WILLIAM, Docent i gronlandsk (eskimoisk) Sprog og Kultur ved Kobenhayns Universitet. BECKETT, Francis, Dr. phil., Docent i Kunsthistorie ved Københavns Universitet, Direktor for den kgl. Afstobningssamling, Ridder af den græske Frelserens Orden. PEDERSEN, JOHANNES PEDER EJLER, Dr. phil. & theol., Professor i semitisk-osterlandsk Filologi ved Kobenhavns Universitet. 11* UDENLANDSKE MEDLEMMER SIEVERS, EDUARD, Dr. phil., Professor i germansk Filologi ved Universitetet i Leipzig. LEFFLER, GOsta MıTTAG-, Dr. phil., fh. Professor i Mathematik ved Højskolen i Stockholm, Kommandør” af Dannebrog og dekoreret med Fortjenstmedaillen i Guld med Krone. Sars, GEORG Ossian, fh. Professor i Zoologi ved Universitetet i Kristiania. BREFELD, Oscar, Dr. phil., Professor i Botanik og Direktor for det botaniske Institut i Breslau. TEGNÉR, Esaras HENRIK VILHELM, Dr. phil. & theol., fh. Professor i østerlandske Sprog ved Universitetet i Lund. BROGGER, VALDEMAR CHRISTOFER, fh. Professor i Mineralogi og Geologi ved Universi- tetet i Kristiania, Ridder af Dannebrog. HAMMARSTEN, OLor, Dr. med. & phil., fh. Professor i medicinsk og fysiologisk Kemi ved Universitetet i Upsala. KLEIN, FELIX, Dr. phil., Professor i Mathematik ved Universitetet i Göttingen. COMPARETTI, Domenico, fh. Professor i Græsk, Firenze. DÖRPFELD, WILHELM, Professor, Dr. phil., fh. forste Sekretær ved det tyske arkæo- logiske Institut i Athen. WiLAMOWITZ-MOELLENDORFF, ULRICH von, Dr. phil., Professor i klassisk Filologi ved Universitetet i Berlin. Picanp, CHARLES-ÉmiLe, Medlem af det franske Institut, Professor i hojere Algebra ved la Faculté des Sciences, Paris. VRIES, HuGo DE, Dr. phil., fh. Professor i Botanik ved Universitetet i Amsterdam. PETTERSSON, OTTO, Dr. phil., fh. Professor i Kemi ved Stockholms Højskole, Kom- mander! af Dannebrog. ENGLER, Aporpn, Dr. phil., Professor i Botanik ved Universitetet i Berlin. GOEBEL, KARL, Dr. phil., Professor i Botanik ved Universitetet i München. PavLov, Ivan Perrovic, Professor i Fysiologi ved det militærmedicinske Akademi i Petrograd. ARRHENIUS, SVANTE, Dr. phil., fh. Professor i Fysik ved Hojskolen i Stockholm, Direk- tor for Nobelinstitutets Afdeling for fysikalsk Kemi, Kommandør” af Dannebrog. Kock, AxEL, Dr. phil., fh. Professor i nordiske Sprog ved Universitetet i Lund, Kommandør” af Dannebrog. NOREEN, ÅDOLF GOTTHARDT, Dr. phil., fh. Professor i de nordiske Sprog ved Uni- versitetet i Upsala. MEYER, Epvuarp, Dr. phil., Professor i Historie ved Universitetet i Berlin. HILDEBRANDSSON, H. H., fh. Professor i Meteorologi og Geografi ved Universitetet i Upsala, Kommandør" af Dannebrog. VocrT, J. H. L., Professor i Metallurgi ved Universitetet i Kristiania. THÉEL, HJALMAR, Dr. phil., fh. Professor og Intendant ved Riksmuseets Evertebrat- afdeling i Stockholm. HıLBERT, Davin, Dr. phil., Professor i Mathematik ved Universiletet i Göttingen. OsTWALb, FRIEDRICH WILHELM, Dr. phil., fh. Professor i Kemi ved Universitetet i Leipzig; Grossbothen, Sachsen. Amira, Kant KoNRAD Ferp. Maria v., Dr. phil., Professor i tysk Ret og Retshistorie ved Universitetet i München. WIDMAN, Oskar, Dr. phil., fh. Professor i Kemi ved Universitetet i Upsala. PENCK, ALBRECHT, Dr. phil., Professor i Geografi ved Universitetet i Berlin. OMONT, HENRI-AUGUSTE, Medlem af det franske Institut, Konservator ved Manu- skript-Departementet i Bibliothèque Nationale i Paris. Eriksson, JAKOB, Dr. phil., fh. Professor, Forstander for den plantefysiologiske og landbrugsbotaniske Afdeling af Landbruks-Akademiens Experimentalfalt ved Stockholm. HiORTDAHL, THORSTEIN HALLAGER, Dr. phil., Professor i Kemi ved Universitetet i Kristiania. LANGLEY, J. N., Dr., Professor i Fysiologi ved Universitetet i Cambridge (England). SCHÜCK, J. Henrik E., Dr. phil., fh. Professor i Æsthetik samt Litteratur- og Kunst- historie ved Universitetet i Upsala. TARANGER, ABSALON, Dr. jur., Professor i Retsvidenskab ved Universitetet i Kristiania. VINOGRADOF, PAuL, Corpus Professor i Retsvidenskab ved Universitetet i Oxford. DREYER, GEORGES, Dr. med., Professor i Pathologi ved Universitetet i Oxford. KossEL, ALBRECHT, Dr. med., Professor i Fysiologi ved Universitetet i Heidelberg. CEDERSCHIOLD, GUSTAF, Dr. phil., fh. Professor i de nordiske Sprog ved Göteborgs Hojskole. Erman, ApoLr, Dr. phil., Professor i Ægyptologi ved Universitetet og Direktor for det Ægyptiske Museum i Berlin. = SRV GEIKIE, Sir ARCHIBALD, Geolog og Mineralog i London. BERTRAND, GABRIEL, Professor i biologisk Kemi ved Sorbonne og Direktor for det biologiske Laboratorium ved Institut Pasteur i Paris. HALLER, ALBIN, Medlem af det franske Institut, Professor i organisk Kemi ved Sorbonne i Paris. Nernst, WALTER, Dr. phil., Professor i fysisk Kemi og Direktor for det fysisk-ke- miske Institut ved Universitetet i Berlin. GRIFFITH, FRANCIS LLEWELLYN, Reader i Ægyptologi ved Universitetet i Oxford. Hunt, ARTHUR SuRRIDGE, Dr., Professor i Papyrologi ved Universitetet i Oxford. Scott, DuNKINFIELD HENRY, fh. Honorary Keeper of the Jodrell Laboratory, Royal Botanic Gardens, Kew, President for Linnean Society of London og for Micro- scopical Society of London. WARBURG, Emit, Dr. phil., Professor, Præsident for den fysisk-tekniske Rigsanstalt, Charlottenburg, Berlin. BEDIER, JOSEPH, Professor i fransk Sprog og Litteratur ved Collège de France, Paris. BERGSON, Henri, Medlem af det franske Akademi, Professor i Filosofi ved Collège de France, Paris. Cumont, FRANZ, Dr. phil., Religionshistoriker, Paris. SCHÄFER, DIETRICH, Dr. phil., Professor i Historie ved Universitetet i Berlin, Kom- mandør” af Dannebrog. WARD, JAMES, Professor i Filosofi ved Universitetet i Cambridge, England. HADAMARD, JACQUES, Medlem af det franske Institut, Professor i Mekanik ved Col- lége de France, og i mathematisk Analyse ved École polytechnique, Paris. MACDONELL, A. A., Professor i Sanskrit ved Universitetet i Oxford. SCHUCHARDT, H., Dr. phil., fh. Professor i romanske Sprog ved Universitetet i Graz. ScHWARTZ, E., Dr. phil., Professor i klassisk Filologi ved Universitetet i Miinchen. SETALA, E. N., Dr. phil., fh. Senator, Professor i finsk Sprog og Litteratur ved Uni- versitetet i Helsingfors. Lorentz, H. A., Dr. phil., Professor i Fysik ved Universitetet i Leiden og Kurator for det fysiske Laboratorium ved Teylers Stiftelse i Harlem. SHERRINGTON, CHARLES S., Professor i Fysiologi ved Universitetet i Oxford. Hacstap, Marius, Professor i det norske Landsmaal og dets Dialekter ved Uni- versitetet i Kristiania. NILSSON, Martin P., Dr. phil., Professor i klassisk fornkunskap og antikens historia ved Universitetet i Lund. OLSEN, MAGNUS BERNHARD, Dr. phil., Professor i oldnorsk og islandsk Sprog og Litteratur ved Universitetet i Kristiania. FALK, HJALMAR S., Dr. phil., Professor i germansk Filologi ved Universitetet i Kri- stiania, LUNDELL, J. A., Dr. phil., fh. Professor i de slaviske Sprog ved Universitetet i Upsala. DANIELSSON, OLOF AuGust, Dr. phil., Professor i græsk Sprog og Litteratur ved Universitetet i Upsala. DE Geer, GERARD Jacos, Friherre, Dr. phil., Professor i Geologi ved Højskolen i Stockholm. ANDREAS, FRIEDRICH CHRISTIAN, Dr. phil., Professor i orientalsk Filologi ved Uni- versitetet i Göttingen. Arnım, Hans v., Dr. phil., Professor i klassisk Filologi ved Universitetet i Wien. BRUNOT, FERDINAND-EUGÈNE, Professor i fransk Sprog ved Sorbonne i Paris. GRIERSON, Sir GEORGE A., M. A., Dr. phil., Orientalist. Hanoraux, GABRIEL, fh. fransk Udenrigsminister, Medlem af det franske Akademi, Historiker. JANET, PIERRE, Professor i Psykologi ved Collège de France i Paris. Linpsay, WALLACE M., Professor i klassisk Filologi ved Universitetet i St. Andrews. MAIER, HEINRICH, Dr. phil., Professor i Filosofi ved Universitetet i Heidelberg. MEILLET, ANTOINE, Professor i sammenlignende Sprogvidenskab ved Collége de France i Paris. PIRENNE, HENRI, Professor i Historie ved Universitetet i Gand. Rasjna, Pio, Professor i romansk Filogi ved Reale Istituto di Studi superiori i Firenze. THOMAS, ANTOINE, Professor i romansk Filologi ved Sorbonne og Directeur d’études ved Ecole pratique des Hautes Etudes i Paris, Medlem af det franske Institut. VITELLI, GIROLAMO, Professor i klassisk Filologi ved Reale Istituto di Studi supe- riori i Firenze. BIRKHOFF, GEORGE A., Professor i Mathematik ved Harvard University, Cambridge, Mass. BORDET, JuLEs, Dr. phil., Direktor for Institut Pasteur i Bruxelles. CuURIE, MARIE, f. SKLODOWSKA, Dr. phil., Professor i Fysik ved Sorbonne i Paris. EINSTEIN, ALBERT, Dr. phil., Professor i theoretisk Fysik ved Kaiser-Wilhelm-Insti- tutet i Berlin. KAMERLINGH ONNES, HEIKE, Dr. phil., Professor i Fysik ved Universitetet i Leiden. Lacroix, ALFRED, Professor i Mineralogi ved Museum d'histoire naturelle i Paris. LEBESGUE, HENRI, Professor i Mathematik ved Collège de France i Paris. Meyer, Hans, Dr. med., Professor i Farmakologi ved Universitetet i Wien. PLanck, Max, Dr. phil., Professor i theoretisk Fysik ved Universitetet i Berlin. RICHARDS, THEODORE WILL., Professor i Kemi ved Harvard University, Cambridge, Mass. Roux, ÉmiLe, Dr. phil., Direktor for Institut Pasteur i Paris, Storkors af Dannebrog. XVI RUTHERFORD, Sir ERNEST, Professor i Fysik ved Universitetet i Cambridge, England. Tuomson, Sir JosEPH JOHN, fh. Professor i Fysik ved Universitetet i Cambridge, England. WIEDEMANN, E., Dr. phil., Professor i Fysik ved Universitetet i Erlangen. WILLSTÄTTER, RicHArp M., Dr. phil., Professor i Kemi ved Uhiversitetet i München. BRADLEY, Francis HERBERT, Dr., Fellow of Merton College i Oxford. Cuny, ALBERT, Professor i latinsk Sprog og sammenlignende Sprogvidenskab ved Universitetet i Bordeaux. Evans, Sir ARTHUR J., Professor i Arkæologi ved Universitetet i Oxford. MEINECKE, FRIEDRICH, Dr. phil., Professor i Historie ved Universitetet i Berlin. MÜLLER, FRIEDRICH WILHELM Kart, Professor, Dr. phil., Afdelingsdirektor ved Museum für Völkerkunde i Berlin. POTTIER, EpMoNnpb, Konservator ved Museum du Louvre i Paris. LINDELOF, ERNST, Dr. phil., Professor i Mathematik ved Universitetet i Helsingfors. Nırsson-EHLE, HERMAN, Dr. phil. & med., Professor i Arvelighedslere ved Univer- sitetet 1 Lund. ALLEN, EDGAR JOHNSON, Dr. Direkter for Marine Biological Association’s Laborato- rium i Plymouth. BARCROFT, JOSEPH, Lecturer i Fysiologi ved Universitetet i Cambridge, England. BATESON, WILLIAM, Professor, Direkter for »John Innes Horticultural Institution«, Merton Park, London. Bay iss, WırLıam Mappock, Professor i almindelig Fysiologi ved University College i London. BENERINCK, MARTINUS WILLEM, Dr., Professor i Mikrobiologi ved den tekniske Hoj- skole i Delft. BRAGG, Sir WırLıam Henry, Professor i Fysik ved University College i London. Davis, Witt1AM Morris, Professor i Geologi og Geografi ved Harvard University, Cambridge, Mass. LANDAU, EpmuNp, Dr. phil., Professor i Mathematik ved Universitetet i Göttingen. MORGAN, Tuomas Hunt, Dr., Professor i Zoologi ved Columbia University i New York. MORGENROTH, JuLIUS, Professor, Dr., Afdelingsdirektor ved Institut »Robert Koch« i Berlin. SMITH, THEoBALD, Dr, Chef for Department of Animal Pathology ved Rockefeller Institute of Medical Research, Princeton, U. S. A. l'HAXTER, ROLAND, Dr., fh. Professor i Botanik ved Harvard University, Cambridge, Mass. XVII VOLTERRA, Vitro, Senator, Professor i theoretisk Astronomi og mathematisk Fysik ved Universitetet i Rom. WASSERMANN, AUGUST v., Professor, Dr., Leder af Kaiser-Wilhelm-Institut für expe- rimentelle Therapie ved Berlin. WEBER, Max, Professor, Direktor for Zoologisk Museum i Amsterdam. WINKLER, Hans, Professor, Direktor for Botanisk Have i Hamburg. Kout, Harvnan, Dr. phil., Professor i Historie ved Universitetet i Kristiania. HEUSLER, ANDREAS, Dr. phil. & jur., Professor i germansk Filologi ved Universitetet i Basel. Snouck HURGRONJE, CHRISTIAAN, Dr., Professor i arabisk Sprog og Islam ved Uni- versitetet i Leiden. APPELL, PAUL, Professor i analytisk og kosmisk Mekanik ved Sorbonne i Paris. HERTWIG, RICHARD, Dr., Professor i Zoologi ved Universitetet i München. LE Cog, ALBERT v., Professor, Dr. phil., Kustos ved Museum für Völkerkunde i Berlin. CHARLIER, CARL VILHELM LupviG, Professor i Astronomi ved Universitetet i Lund. FORSSMAN, MAGNUS JOHN CARL AuGust, Dr. med., Professor i almindelig Pathologi ved Universitetet i Lund. FÜRST, Cart MAGNUS, fh. Professor i Anatomi ved Universitetet i Lund, Ridder af Dannebrog. BATHER, Francis ARTHUR, Dr., Keeper ved British Museum (Natural History) i London. Bower, FREDERICK ORPEN, Professor i Botanik ved Universitetet i Glasgow. BROWNE, EDWARD GRANVILLE, Professor i Arabisk ved Universitetet i Cambridge. GARDINER, ALAN HENDERSON, Dr., Ægyptolog, London. JAEGER, WERNER W., Professor i klassisk Filologi ved Universitetet i Berlin. STERN, ALFRED, Professor i almindelig Historie ved Eidgenôssische Technische Hoch- schule i Zürich. ; Harpy, GODFRED HAROLD, Professor i Mathematik ved Universitetet i Oxford. III tog | ar Fr ATU TT ef In! v3 if i 10-29 " ns "IU ENE et LT Bi hay } for ; | ’ FEN?" hie OF K ANN f 71 ' ts ‘CPA Ki « re le : J a" NA ee Beer » AU) weit ‘] : LI OR i , £ P ' i rr a “ae i 7 v må R a i SATS U i if Å > i iv ey - \ I NE: f nt ‘ie % y “Gl i H 3 UL 3: HEN å "sg D {ae son RARE EL: à i <= K i rå 8 A DIE RHODANIDE DES GOLDES UND DAS FREIE RHODAN MIT EINEM ANHANG UBER DAS GOLDCHLORID VON NIELS BJERRUM uno AAGE KIRSCHNER D. KGL. DANSKE VIDENSK. SELSK. SKRIFTER, NATURVIDENSK. OG MATHEM. AFD., 8. RÆKKE, V. 1 KOBENHAVN HOVEDKOMMISSIONÆR: ANDR. FRED. HOST & SON, KGL. HOF-BOGHANDEL BIANCO LUNOS BOGTRYKKERI 1918 vie 14 OB ats 3 r SHIRT 12 der vorliegenden Arbeit werden die Rhodanide des Goldes untersucht. Diese Ver- bindungen sind alle komplex, und sie verbinden eine grosse Komplexität mit einer kleinen Robustheit, d. h. sie besitzen trotz geringer Abspaltung von Goldionen eine grosse Reak- tionsgeschwindigkeit bei Doppeldekompositionen. Die Goldrhodanide haben uns interes- siert, erstens als Vertreter von stark komplexen, aber nur wenig robusten Salzen und zweitens, weil sie Untersuchungen über die Umwandlungen der verschiedenen Oxyda- tionsstufen eines Metalles ineinander erlaubt haben. Unsere Untersuchungen über das Verhalten des Aurirhodanids haben uns dazu geführt, in diesen Lésungen die Existenz des freien Rhodans anzunehmen, und wir haben eine Reihe Eigenschaften dieses freien Rhodans bestimmt. In dem ersten Abschnitt dieser Arbeit wird eine Ubersicht über den Gang und die Resultate unserer Untersuchungen gegeben, und in den folgenden Abschnitten werden die experimentellen Einzelheiten ausführlicher mitgeteilt. I. Ubersicht. Feste goldrhodanidhaltige Verbindungen. Die Goldrhodanide sind bisher nur wenig untersucht worden, was um so merkwiirdiger ist, als sie in verschiedenen Weisen tech- nische Anwendung gefunden haben, z. B. zur Ténung von photographischen Kopien. Die einzige grössere Arbeit über die Goldrhodanide ist die 1863 von P. T. CLEvE*) publizierte. CLEVE hat sich hauptsächlich mit der Darstellung von festen Verbindungen be- schäftigt. Nach ihm ist es nicht möglich, die einfachen Goldrhodanide selbst in fester Form darzustellen. Dagegen kann man leicht aus Kaliumrhodanid und Goldchlorid ein Doppelsalz mit der Zusammensetzung, Au Rh,, K Rh, gewinnen. Dieses Kaliumauri- rhodanid ist ziegelrot und schwerlôslich in kaltem Wasser; in heissem Wasser wird es leicht aufgenommen, aber nur unter gleichzeitiger Zersetzung; denn durch Eindampfen der gelben Lösung erhält man nicht Kaliumaurirhodanid, sondern ein unreines Gemenge, in welchem es CLEVE gelang, ein leichtlésliches Doppelsalz von Aurorhodanid und Ka- liumrhodanid, das Kaliumaurorhodanid, AuRh, K Rh, nachzuweisen. 1) Ofversigt af Kgl. Vet.-Akad, Förh. 20, 233 (1863). Auszug in: Journ, f. prakt. Chem. 94, 14 (1865). 1* Im Laufe unserer Untersuchungen haben wir sowohl das CLeve’sche Kaliumauri- rhodanid wie analog zusammengesetzte Doppelsalze von Ammonium und Natrium dar- gestellt und analysiert. Diese Salze werden leicht aus den entsprechenden Rhodaniden durch Zusatz von Aurichlorid gewonnen, wenn man nur mit Uberschuss von Rhodanid und in saurer Lésung arbeitet, um Zersetzung zu verhindern. Das Ammoniumauri- rhodanid ist ebenso schwerlôslich wie das Kaliumsalz; aber das Natriumaurirho- danid ist bedeutend leichter lôslich, wenn auch nicht leicht lôslich. Für das Kaliumsalz ist das Léslichkeitsprodukt ca. 8-10? und für das Natriumsalz ca. 2:107%, Eine ähn- liche Löslichkeit wie das Natriumsalz besitzt das Baryumsalz, das wir jedoch nicht analysiert haben. Auch das entsprechende Wasserstoffaurirhodanid, H Rh, Au Rh;, 2 H,0, haben wir in schôn krystallinischer Form gewonnen. Es krystallisiert mit zwei Molekülen Wasser. Man kann diese Säure aus einer salzsauren Lösung von Natriumrho- danid und Aurichlorid mit Ather ausschütteln und durch Eindampfen der ätherischen Schicht in krystallinischer Form gewinnen. Für das CLeve’sche Kaliumaurorhodanid haben wir eine neue und bessere Dar- stellungsweise gefunden. Aus einer salzsauren Lösung von Natriumrhodanid und Auri- chlorid wird durch Reduktion mit Natriumsulfit eine aurorhodanidhaltige Lösung ge- wonnen; hieraus wird mit Ather das Wasserstoffaurorhodanid ausgeschiittelt, und durch Neutralisation der ätherischen Lösung mit Kaliumbikarbonat wird das Kaliumaurorho- danid gewonnen. In dieser Weise dargestellt ist es schneeweiss, wahrend CLEVE es als strohgelb bezeichnet; wir schliessen hieraus, dass er es nicht rein in Händen gehabt hat. Durch Neutralisation der ätherischen Schicht mit Ammoniak haben wir das entsprechen- de Ammoniumaurorhodanid erhalten. Aus der atherischen Lésung von Wasserstoffaurorhodanid wird durch langeres Be- handeln mit Uberschuss von Ammoniak das schon von CLEVE dargestellte Monammino- aurorhodanid, AuNH,Rh, in Form eines schneeweissen Niederschlags gewonnen. In dieser Weise kann man diese Verbindung leicht rein und in grosser Menge erhalten, wäh- rend CLEVE sie nur in ganz kleiner Menge dargestellt hat. Darstellung von goldrhodanidhaltigen Lösungen. Wenn man Natriumrhodanid in verdünnter Salzsäure löst und Aurichlorid in geringer Menge zufiigt, so erhält man eine rotbraune Lösung, in welcher das Gold als ein rotbrauner Aurirhodanidkom- plex vorhanden ist. Die rote Farbe stellt sich augenblicklich mit voller Stärke ein, und man kann daraus schliessen, dass die Rhodanionen momentan Chlorionen aus dem Auri- chloridkomplex austreiben. Zur Darstellung von aurirhodanidhaltigen Lösungen sind Kaliumrhodanid und Ammoniumrhodanid wegen der Schwerlöslichkeit von Kalium- und Ammoniumaurirhodanid nicht so geeignet wie Natriumrhodanid und Baryumrhodanid. Die rotbraunen Aurirhodanidlösungen können durch vorsichtigen Zusatz von Na- triumsulfit entfärbt werden; hierdurch wird der rote Aurirhodanidkomplex zu einem farb- losen Aurorhodanidkomplex reduziert, und man erhält in dieser Weise leicht aurorho- danidhaltige Lösungen. Die Entfärbung geht so glatt und schnell vor sich, dass man mit guter Genauigkeit die Menge von Aurigold in einer rhodanidhaltigen Lösung durch Titra- tion mit einer verdünnten Lösung von Natriumsulfit bis zur Entfärbung bestimmen kann (Sulfittitrierung). Zusammensetzung und Komplexität der farblosen Aurorhodanidkomplexe. Um den - J Zustand des Goldes in den farblosen Aurorhodanidlösungen zu bestimmen, haben wir das elektrische Potential einer Goldelektrode, die in eine solche Lösung eintaucht, gemessen. Es zeigte sich dabei, dass dieses Potential mit der Aurogoldkonzentration, [Au], und der Rhodanionenkonzentration, [Rh~], nach der folgenden Formel variiert: år) En 0,689 — 0,058 log [Rh]? Das Potential ist hier in Volt angegeben, und der Index À deutet an, dass das Poten- tial gegen eine Normal-Wasserstoffelektrode gemessen wird. Aus der Form des zweiten Gliedes kann man nach BopLANDER!) schliessen, dass das Gold in der L.ösung als Dirho- danoauroat-Ion, Au Rh,, vorhanden ist. Die Potentialmessungen geben uns aber nicht nur die Zusammensetzung, sondern auch die Komplexität des Aurorhodanidkomplexes; denn der Zahlenwert des ersten Gliedes, 0,689, das sogenannte NormalpotentialGold- Aurorhodanid, ist ein Mass der Affinität zwischen Gold und Rhodan in dem Komplex. Je kleiner das Normalpotential für einen Aurokomplex ist, um so grösser ist seine Komplexität. Um aus dem Normalpotential die Komplexitätskonstante des Aurorhodanids, Pe _ [Au Rh, ] [Au*]- [Rh]? berechnen zu können, muss man das Normalpotential Gold-Auroion kennen. Nach ABEGG und CAMPBELL?) ist dieses Normalpotential ca. 1,5. Sie haben gefunden, dass Aurooxyd in konzentrierter Salpetersäure ein wenig löslich ist, und haben das Potential einer Gold- elektrode in solchen Lösungen gemessen. Aus ihren Messungen haben sie den Wert des Nor- malpotentials Gold-Auroion berechnet, unter der Annahme, dass alles Gold in der Lösung als Auroion vorhanden ist, und dass der potentialbestimmende Vorgang die Entladung dieser Auroionen ist. Schon die erste Annahme ist wohl sehr zweifelhaft, und was die zweite betrifft, so haben wir für eine platinierte Platinelektrode in goldfreier Salpeter- säure ähnliche Potentialwerte gefunden, wie ABEGG und CAMPBELL für eine Goldelektrode in der goldhaltigen Salpetersäure. In Tabelle 1 sind unsere Messungen mit denjenigen von ÄBEGG und CAMPBELL zusammengestellt. Tabelle 1. Potential von Pf oder Au | HNO, | gesätt. NH,NO, 1m KCl, HgCl, Hg bei 25°. Au in HNO, mit à i ; Konz. von HNO, Au, O gesattigt. D TT HNO, FAP K. u. B. 11,66 n 0,998 Volt 0,973 7,76 n 0,901 - 0,913 3,32 n 0,835 - 0,800 1) DEDEKIND-Festschrift, Braunschweig 1901, 151. Ber. deut. chem. Ges. 36, 3933 (1903). ?) Zeitschr. für Elektrochemie 13, 440 (1907). Trans. Faraday Soc. 3, Mai 1907. Wir schliessen aus der annähernden Ubereinstimmung zwischen den beiden verschie- denen Messungsreihen, dass das von ABEGG und CAMPBELL gemessene Potential einfach das Oxydationspotential der Salpetersäure ist. Hiernach ist das Normalpotential Gold-Auroion noch nicht bekannt, und wir kénnen also aus dem Normalpotential Gold-Aurorhodanid die Komplexitätskonstante des Aurorhodanids noch nicht berechnen. Mit Hilfe des Normalpotentials Gold-Aurorhodanid kann man aber trotzdem die Komplexität des Aurorhodanids mit denen anderer Aurokomplexe vergleichen, fiir wel- che die Normalpotentiale gegen Gold bekannt sind. Mit Hilfe des von BopLANDER gemes- senen Normalpotentials Gold-Aurocyanid, — 0,611, haben wir z. B. berechnet, dass die Komplexitätskonstante des Dirhodanoauroatkomplexes 102%*mal kleiner ist als die des Dicyanoauroatkomplexes. Zusammensetzung und Komplexität der rotbraunen Aurirhodanidkomplexe. Um den Zustand des Goldes in den rotbraunen Aurirhodanidlésungen zu bestimmen, war es notwendig, etwas anders als bei Aurorhodanid zu verfahren. Das Potential Gold- Aurirhodanid ist nämlich nicht direkt messbar, weil Gold in Aurirhodanid unter Bildung von Aurorhodanid gelöst wird. Wir haben das Potential einer Platinelektrode gemessen, die in eine gemischte Lösung von Aurorhodanid und Aurirhodanid ein- taucht. Das Platin wirkt hier als indifferente Elektrode, und man misst das Potential Aurorhodanid-Aurirhodanid. Für dieses Potential wurde gefunden: 0,058; [Au] En — 0,6454 2" log Eau. Rk" Hier bedeutet [Au””] die molare Konzentration des Aurigoldes. Diese Formel zeigt, dass der Aurirhodanidkomplex zwei Rhodangruppen mehr enthalt als der Aurorhodanid- komplex; da der Aurorhodanidkomplex die Formel Au Rh,” besitzt, muss somit der Auri- rhodanidkomplex die Formel Au Rh,” besitzen, und als chemische Gleichung des Ele- trodenvorganges ergibt sich: Au Rh,+28 = AuRh, + 2Rh, wenn man ein negatives Elektron mit 6 bezeichnet. Die Grésse 0,645 wird das Normalpotential Aurorhodanid- ARERR ge- nannt. Um für die Komplexität des Aurirhodanids ein Mass zu erhalten, muss man das Normalpotential Gold-Aurirhodanid kennen. Dieses Potential ist, wie oben bemerkt wurde, nicht direkt messbar; man kann es aber nach einem Verfahren von LUTHER!) aus den Normalpotentialen Gold-Aurorhodanid (0,689) und Aurorhodanid-Aurirhodanid (0,645) berechnen, und wir haben in dieser Weise den Wert 0,660 gefunden. 3 En — 0,689 + 2. 0.645; En — 0,660. Diese Berechnungsweise beruht darauf, dass die elektrische Energie, die zur Bildung von Aurirhodanid aus Gold und Rhodanionen notwendig ist, dieselbe sein muss, sei es dass das Aurirhodanid iiber die Zwischenstufe Aurorhodanid oder direkt gebildet wird. ') Zeitschr. physik. Chem. 36, 385 (1901). Aus dem Normalpotential Gold-Aurirhodanid ist es nicht möglich, die Komplexi tatskonstante des Aurirhodanidkomplexes, pr JE. [Au***] x [Rh r ’ zu berechnen, da das Normalpotential Gold-Auriion zur Zeit nicht bekannt ist. Wir können aber mit seiner Hilfe die Komplexität des Aurirhodanids mit der anderer Auri-, komplexe vergleichen. Um einen solchen Vergleich ausführen zu können, haben wir das Normalpotential Gold-Aurichlorid gemessen); es beträgt 1,001. Hieraus berechnet man, dass die Komplexitätskonstante des Aurirhodanids 10'”’mal grösser ist als die des Aurichlorids. Aus unseren Potentialmessungen können wir schliessen, dass das Aurirhodanid in den angewandten Lösungen als Tetrarhodanoauriation vorhanden ist. In diesen Lösungen war die Rhodanionenkonzentration 0,1 bis 0,4 molar und die Wasserstoffionenkonzen- tration ca. 1 molar. Durch spektrophotometrische Messungen und Löslichkeitsbestim- mungen von Natriumaurirhodanid haben wir gefunden, dass bei höheren Rhodanionen- konzentrationen rhodanreichere Komplexe gebildet werden, und dass diese bei einer Rhodanionenkonzentration von 2 bis 4 molar überwiegen. Aus spektrophotometrischen Messungen von Lösungen mit kleinerer Rhodanionenkonzentration und Wasserstoflio- nenkonzentration glauben wir weiter schliessen zu dürfen, dass der Tetrarhodanoauriat- komplex in solchen Lösungen unter Abspaltung von Salzsäure und Bildung von hy- droxylhaltigen Komplexen hydrolysiert wird, z, B.: i Au Rh, + H,0 = AuRh,(OH)~ + H+ + Rh” Au Rh, (OH) + H,O = AuRh,(OH);+ H*+ Rh. Diese Hydrolyse scheint doch erst gross zu werden, wenn das Produkt der Wasser- stoflionenkonzentration und der Rhodanionenkonzentration kleiner als 0,01 wird. Für den Dirhodanoauroatkomplex tritt eine ähnliche Hydrolyse erst viel später ein; in rho- danionenreicher Lösung ist sie noch beim Umschlagspunkt des Phenolphtaléins nicht merkbar; denn man kann in Gegenwart von reichlichem Rhodanid den Säuregehalt einer Aurorhodanidlösung mit Phenolphtaléin als Indikator titrieren. Über die Unbeständigkeit der Goldrhodanide. Alle goldrhodanidhaltigen Lösungen sind unbeständig. Sie scheiden je nach den äusseren Umständen nach kürzerer oder längerer Zeit das Gold in metallischer Form aus. Wir haben gefunden, dass bei dieser Zersetzung hauptsächlich zwei Vorgänge wirksam sind, erstens eine nicht umkehrbare Autoreduktion nach dem Schema: | 3AuRh, + 4H,0 > 3AuRh; +5HRh+ HCN + H,SO, und zweitens eine umkehrbare Goldausscheidung aus Aurorhodanid unter gleich- zeitiger Bildung von Aurirhodanid: | 3AuRh; 2 2Au + AuRhz + 2Rh-. 1) Vgl. den Anhang dieser Abhandlung. 8 Durch das Zusammenspiel dieser beiden Vorgange entstehen die oft recht verwickelten Verhältnisse, die sich bei der Zersetzung der Goldrhodanide zeigen. Wir haben diese beiden Vorgänge einer ausführlichen Untersuchung unterworfen. Die Autoreduktion. In verdünnter Lösung bei gewöhnlicher Temperatur geht die Goldausscheidung so langsam vor sich, dass man die Autoreduktion des Aurirhoda- nids untersuchen kann, ohne durch Goldausscheidung gestört zu werden. Wir haben deshalb die oben angegebene Gleichung für die Autoreduk- tion durch gewichtanalytische Schwefelsäurebestimmungen und durch massanalytische Säurebestimmungen verifizieren können. i Den zeitlichen Verlauf der Autoreduktion kann man ver- folgen, indem man beobachtet, wie die rotbraune Farbe des Aurirhodanids nach und nach schwåcher wird. Besser ist es aber, das noch vorhandene Auri- rhodanid durch Sulfittitrierung zu bestimmen, und in dieser 100 % 80% 60% 40% 20% Weise haben wir die meisten S unserer Messungen ausgeführt. 3 In Fig. 1 ist als Beispiel der = Verlauf in 0,004 m HAuCl, 0 Minuten 40 60 120 160 200 0,3 m NaRh, 0,1 m HCl 1) auf- Fig. 1. Die Autoreduktion in 0,004m HAuCl,, 0,3m NaRh, gezeichnet; die einzelnen experi- 0,1 m HCI bei 18°. mentellen Bestimmungen sind durch kleine Kreuze markiert. Der Reaktionsverlauf stimmt gar nicht mit den gewöhnlichen Verläufen fir mono-, di- und trimolekulare Reaktionen, die in der Figur durch die Kurven I, II und III angegeben sind. Fir die Autoreduktion nimmt die Geschwindigkeit mit der Zeit viel schneller ab als fiir mono- und di-, ja selbst trimolekulare Reaktionen. Wir entdeckten bald, dass diese schnelle Geschwindigkeitsabnahme damit in Verbindung stand, dass das gebildete Aurorhodanid hemmend wirkte; denn wir konnten durch Zusatz von Aurorhodanid die Geschwindigkeit schon von Anfang an weit hinunterdriicken. Eine solche Hemmung wirde leicht verstandlich sein, wenn der Vorgang umkehrbar ware; da der Vorgang aber sicher einseitig verlauft, indem Schwefelsäure und Cyanwasserstoff nicht Aurorhodanid zu Aurirhodanid oxydieren kénnen, muss man hierfür eine besondere Erklärung finden. ) Die Zusammensetzung von gemischten Lösungen wird durch die Konzentrationen ange- geben, die die zusammengemischten Körper in der Lösung besitzen würden, wenn sie nicht mitein- ander reagierten. Lange gelang es trotz vieler Bemühungen nicht, eine befriedigende Erklärung zu finden, ja es gelang nicht einmal, eine empirische Formel aufzustellen, die den zeitlichen Verlauf in allen untersuchten Lösungen darstellen konnte. Erst nachdem wir durch fortgesetzte Untersuchungen entdeckt hatten, dass der Aurirhodanidkomplex nach der Gleichung Au Rhy = AnRhy + Rh, etwas dissoziiert ist, gelang es, das Rätsel zu lösen. Die Dissoziation des Aurirhodanids zu Aurorhodanid und Rhodan. Das erste An- zeichen einer Dissoziation von Aurirhodanid unter Bildung von Aurorhodanid wurde durch einige Potentialmessungen von Platinelektroden in frischen Aurirho- danidlösungen erhalten. Aus der Messung des Potentials einer Platinelektrode kann man nach der Formel auf Seite 6 das Verhältnis zwischen Aurorhodanid und Aurirho- danid in der Lösung berechnen. Unsere Potentialmessungen zeigten nun, dass in den untersuchten frischen Aurirhodanidlösungen 10 bis 20°/, des Goldes als Aurorhodanid vorhanden waren. Da in denselben ‚Lösungen, nach Sulfittitrierungen zu urteilen, alles Gold als Aurirhodanid vorhanden war, wurden wir zu der Annahme geführt, dass das elektrometrisch nachweisbare Aurorhodanid durch eine umkehrbare Dis- soziation aus dem Aurirhodanid gebildet sein musste; denn nur dann bleibt das elektrometrisch nachweisbare Aurorhodanid bei der Sulfittitrierung unentdeckt. Diese Dissoziationshypothese haben wir durch spektrophotometrische Messungen bestätigen können. Nach der Hypothese muss. die Farbe einer Aurirhodanidlösung durch Zusatz von Aurorhodanid stärker werden; denn das Aurorhodanid muss die Dis- soziation des Aurirhodanids zurückdrängen. Wir haben diese Zunahme der Farbstärke festgestellt und haben aus ihr ähnliche Werte für die Dissoziation des Aurirhodanids berechnet wie nach den Potentialmessungen. Tabelle 2 enthält die nach den zwei ver- schiedenen Verfahren berechneten Dissoziationsgrade. Tabelte 2. Dissoziation von Aurirhodanid nach der Gleichung AuRh, — AuRh, + Rh,. mol. spektro- ” elektrometrisch Konz. photometrisch I II 0,001 — 16,1 °/o — 0,002 14,4% 12,5 lo 17,6 lo 0,004 — 9,9 %o 11,2 %o 0,005 9,5 lo — — Unsere Versuche ergeben als unmittelbares Resultat, dass das eine Dissoziations- produkt von dem Tetrarhodanoauriation das Dirhodanoauroation ist. Das zweite Disso- ziationsprodukt muss die Zusammensetzung Rh, besitzen und muss also das bisher un- bekannte freie Rhodan sein!). Wir haben diesen Schluss daraus gezogen, dass der 1) Nach Abschluss dieser Arbeit, deren Resultate der königl. dänischen Gesellschaft der Wissen- schaften in der Sitzung 8. Febr. 1918 vorgelegt wurden, ist die Darstellung des freien Rhodans in einer Inaugural-Dissertation, Upsala April 1918, von SÖDERBÄcK beschrieben worden. D.K,D, Vidensk, Selsk. Skr., naturvidensk. og mathem, Afd., 8. Række, V. 1. 2 10 Dissoziationsgrad von der Wasserstoffionenkonzentration und der Rhodanionenkon- zentration unabhangig ist, wie man aus Tabelle 3 ersehen kann. Tabelle 3. Dissoziationsgrad von Aurirhodanid nach der Gleichung: AuRh, — AuRh, + Rh,. 0,002m HAuCl, 0,005m HAuCl, 0,056m NaRh, 1m HCl 14,2 0/0 11,7 %o 014m — Im — 14,4 lo 9,7 “lo 028m — 1m — 14,7 0/0 9,4 0 0,28 m — 05m — — 6,8 Plo 023m — 02m — — 9,8 "lo Daraus folgt nåmlich, dass Wasserstoffionen und Rhodanionen bei der Dissozia- tion weder gebildet noch verbraucht werden. Da weiter der Dissoziationsgrad sich mit der 2 = in ändert, können wir schliessen, dass der Aurirhodanidkomplex, AuRhy, nur in zwei Teile dissoziiert wird; da der eine Teil der Aurorhodanidkomplex, AuRhz, ist, muss der zweite Teil das freie Dirhodan, Rh,, sein (oder möglicherweise ein Hydrat davon). Für die Dissoziationskonstante des Aurirhodanids haben wir als Mittel aus unseren Mes- sungen den Wert 0,49.10”* berechnet. Es gilt also: Aurirhodanidkonzentration nach dem OsiwAp’schen Verdünnungsgesetz, C- [Au Rhz]- [Rhy] x —4 Ws ag 0,49 : 10-4, Die Deutung des Verlaufes der Autoreduktion. Da das freie Rhodan bisher nicht entdeckt worden war, musste es ohne Zweifel recht unbeständig sein; es lag daher nahe, anzunehmen, dass die Autoreduktion des Aurirhodanids durch diese Unbestän- digkeit des freien Rhodans verursacht wurde. Man sieht leicht ein, dass diese Annahme jedenfalls qualitativ den anomalen Reaktionsverlauf der Autoreduktion erklären kann. Wenn die Reaktion fortschreitet, bildet sich Aurorhodanid; hierdurch wird die Disso- ziation des Aurirhodanids zuriickgedrangt, und die Menge des freien Rhodans nimmt stark ab. Wenn also die Autoreduktion über das freie Rhodan verläuft, so wird hiernach die grosse Verminderung der Geschwindigkeit bei fortschreitender Autoreduktion leicht verstandlich, und man versteht weiter auch die hemmende Wirkung eines Zusatzes von Aurorhodanid. Wenn man den Ansatz macht, dass die Zersetzungsgeschwindigkeit des Rhodans der zweiten Potenz der Rhodankonzentration proportional ist, und wenn man die Rhodankonzentration nach der oben aufgestellten Dissoziationsgleichung berechnet, erhält man sogleich die früher vergebens gesuchte Formel des zeitlichen Verlaufs der Autoreduktion. Den Grad der gefundenen Ubereinstimmung kann man aus Figur 1, Seite 8 ersehen. Wenn man für die Geschwindigkeitskonstante bei der Zersetzung des Rhodans den Wert 3600 benutzt, erhalt man fiir den Verlauf der Autoreduktion die 4. Kurve die sich den Messungsresultaten mit genügender Annäherung anschmiegt. Eine ähnliche Ubereinstimmung fanden wir bei der Berechnung unserer anderen Ver- suche, 11 Die Autoreduktion wird von Wasserstoffionen und Rhodanionen sehr stark ge- hemmt, und deshalb ist auch die Geschwindigkeitskonstante, die man fiir die Zersetzung des Rhodans aus unseren Autoreduktionsmessungen berechnet, um so kleiner, je grösser die Wasserstoffionenkonzentration und die Rhodanionenkonzentration sind. Sie ist annähernd den Quadraten von diesen Konzentrationen umgekehrt proportional. Unsere Messungen der Geschwindigkeit der Autoreduktion bei 18° können mit grober Annähe- rung in folgende Formel zusammengefasst werden: d[Au] _, 5 MR a OEP TRE Es ist leicht, eine einfache Erklärung dieser Formel zu finden. Wir brauchen nur anzunehmen, dass das freie Rhodan analog dem Chlor nach der Gleichung: Rh, + H,O = RhOH + Rk + Ht etwas hydrolysiert ist. Die Verbindung RAOH können wir passend unterrhodanige Säure nennen. Wenn dieses Hydrolysengleichgewicht sich momentan einstellt, und die Autoreduktion in der Weise vor sich geht, dass zwei Moleküle unterrhodanige Säure miteinander reagieren, muss die Geschwindigkeit der zweiten Potenz der Konzentra- tion der unterrhodanigen Säure proportional sein: III — MT _ x. 4240 Hp. (1) Nach dem Massenwirkungsgesetz gilt: [RhOH] [Rh] LH _ [Rh] K. (2) Aus (1) und (2) folgt durch Elimination von [RhOH]: d[AUTT] _ , ge E 3 Bas OLES Le d. h. ein Ausdruck von der gewiinschten Form. Nachdem wir durch unsere kinetischen Untersuchungen zu dem Resultate gekom- men sind, dass der geschwindigkeitsbestimmende Vorgang eine Reaktion zwischen zwei Molekiilen unterrhodanige Säure ist, haben wir weiter zu erwägen, was aus diesen zwei Molekülen entstehen kann. Wahrscheinlicherweise wird hierbei rhodanige Säure, HRhO, gebildet: 2HRhO = HRhO, + HF” + Rh. Die rhodanige Säure wird dann augenblicklich nach ihrer Bildung von einem dritten Mole- kil HRAO zu Rhodansäure, HRhO,, oxydiert: HRhO,+ HRhO = HRhO,+ H*+ Rh‘, 12 und Rhodansäure ergibt endlich durch Hydrolyse die beobachtbaren Endprodukte Schwefelsäure und Cyanwasserstoff : HRhO,+ H,0 = HCNSO,+ H,0 = HCN + H,S0,.7) Hiernach haben wir die fiir die Autoreduktion charakteristische Bruttogleichung: 3AuRh, + 4H,0 = 3AuRh;+5H+-+5Rh +HCN+H,SO, in die folgenden fünf Partialgleichungen aufgelöst: AuRh, S AuRh, + Rh, momentan. (1) Rh, + H,O 2 RROH+H++Rh momentan. (2) 2RhOH > HRhO,+H++Rh schnell. (3) H RhO, + HRRO > HRhAO, + H+ + Rh momentan. (4) HRhO, + H,0 > HCN + H,S0, momentan. _ (5) Von diesen Vorgängen sind (1), (2), (4) und (5) momentan verlaufend, und (3) allein bestimmt durch seinen nicht momentanen Verlauf die Geschwindigkeit der Autoreduk- tion. (1) und (2) sind reversibel, und (3), (4) und (5) sind irreversibel. Über das freie Rhodan. Nachdem wir gefunden haben, dass eine Aurirhodanid- lösung freies Rhodan enthält, und dass man zu dem für die Autoreduktion gefundenen Reaktionsverlauf kommt, wenn man einen Verlauf über das freie Rhodan als Zwischen- produkt annimmt, ist nach unserer Meinung diese Annahme von Rhodan als Zwischen- produkt vollständig berechtigt. Um aber dieser Annahme jeden hypothetischen Cha- rakter zu nehmen, haben wir durch direkte Messungen festgestellt, dass die Zersetzung des freien Rhodans wirklich nach den Gesetzen und mit der Geschwindigkeit vor sich geht, die wir aus unseren Autoreduktionsuntersuchungen berechnet haben. Dabei ver- fuhren wir folgendermassen. Aus dem Normalpotential Aurorhodanid-Aurirhodanid, E, und der Dissoziations- konstante des Aurirhodanids, K, durch welche die Konzentration des freien Rhodans in einer Aurorhodanid-Aurirhodanid-Lösung bestimmt ist, kann man das Normalpotential Rhodan-Rhodanion, Æ', berechnen, indem: E = E — —.InK. Nach dieser Formel findet man für das Normalpotential Rhodan-Rhodanion den Wert 0,769; für Jod ist das entsprechende Normalpotential 0,54, und für Brom ist es 1,09. Aus diesen Werten kann man schliessen, dass Rhodan aus einer wässrigen Rhodanionenlösung von Brom freigemacht wird, und dass es selbst Jod aus einer wässrigen Jodionenlösung frei- machen kann. Das freie Rhodan liegt also in der Reihe der freien Halogene zwischen Brom- und Jod. Wenn man Brom zu einer salzsauren Lösung von Natriumrhodanid setzt, verschwindet die Farbe des Broms wirklich’ auch sogleich, und man erhält eine 1) Eine ähnliche Reaktionsfolge hat auch S6pERBACK in seiner Dissertation, Upsala 1918. 13 farblose Lösung, die Jod aus Kaliumjodid freimacht; das braune Brom bildet also augen- blicklich farbloses freies Rhodan, und dies farblose Rhodan macht wieder aus Kalium- jodid braunes Jod frei. Die Fähigkeit der mit Brom versetzten Rhodanidlösung zur Jodausscheidung verschwindet aber schnell, ganz in Übereinstimmung damit, dass das freie Rhodan unbeständig ist und schnell in Schwefelsäure und Cyanwasserstoff um- gewandelt wird. Durch Bestimmung der Geschwindigkeit, mit welcher die Fähigkeit zur Jodausscheidung bei einer mit Brom versetzten salzsauren Natriumrhodanidlösung verschwindet, haben wir gefunden, dass die Zersetzung des Rhodans bimolekular ver- läuft und sowohl dem Quadrat der Wasserstoffionenkonzentration wie dem der Rhodanionenkonzentration proportional ist; und für die Geschwindigkeitskonstante, k, in der Gleichung: __ a[Rh,] he _ [RA,} dl [H+}?- [RA |’ haben wir Werte in der Umgebung von 5 erhalten. Unsere früher aufgestellte Formel für die Zersetzungsgeschwindigkeit des Rhodans wird hierdurch vollständig bestätigt, und unsere Hypothese, wonach die Autoreduktion über das freie Rhodan verläuft, wird zur Gewissheit erhoben. RTE Uber die Berechnung von Oxydationsgeschwindigkeiten aus Oxydationspotentialen. Ohne Zweifel werden viele Oxydationsmittel, wenn sie zur Oxydation von Rhodaniden benutzt werden, nach einem ähnlichen Mechanismus wie das Aurigold wirken. Zuerst wird durch einen schnellen umkehrbaren Vorgang freies Rhodan gebildet, und dieses zersetzt sich dann langsamer. Wahrscheinlicherweise werden Oxydationsmittel wie Jod, Ferrisalze, Kuprisalze Rhodanide in dieser Weise oxydieren. Fiir diese Oxydationsmittel wird es dann möglich sein, die Geschwindigkeit ihrer oxydierenden Wirkung auf Rhodanide aus ihren Oxydationspotentialen zu berechnen. Bekanntlich gehört die Aufgabe, einen Zusammenhang zwischen Affinität und Reaktionsgeschwindigkeit zu finden, zu den nicht gelösten Problemen der Chemie, und wahrscheinlicherweise ist diese Aufgabe überhaupt nicht allgemein zu lösen. Um so grösser ist aber das Interesse, das sich an partielle Lösungen dieser Aufgabe knüpft. Die Goldausscheidung. Das bei der Autoreduktion gebildete Aurorhodanid ist nicht das Endprodukt, sondern spaltet sich langsam in Aurirhodanid und Gold nach der Gleichung: 3AuRh, — 2Au + AuRh, +2Rh. ‘ Unter Bedingungen, wo die Autoreduktion sehr langsam verläuft, haben wir ex- perimentell das von der obenstehenden Gleichung geforderte Verhältnis zwischen dem gebildeten Aurirhodanid und Gold gefunden. Wir werden diese Reaktion kurz als die Goldausscheidung bezeichnen. Die Goldausscheidung ist im Gegensatz zu der Auto- reduktion umkehrbar: von einer reinen Aurirhodanidlésung wird Gold unter Bildung von Aurorhodanid gelöst. Für Goldrhodanidlösungen, die mit metallischem Gold in Gleichgewicht sind, muss nach dem Massenwirkungsgesetz die folgende Gleichung er- füllt sein: | [AuRh,]- [Rh 7? u [Au Rh, ]' 14 Den Wert von K, die Gleichgewichtskonstante, haben wir nicht direkt durch Analyse von Gleichgewichtslösungen bestimmen können; ein vollständiges Erreichen des Gleichgewichts wird nämlich durch die Autoreduktion unmöglich gemacht, da durch sie das Aurirhodanid ständig entfernt wird. Wir haben aber ihren Wert aus den Normal- potentialen Gold-Aurorhodanid (0,645) und Aurorhodanid-Aurirhodanid (0,689) zu 33 berechnet. Denn INK = 0,689 — 0,645 gibt K — 33, Die Gleichgewichtsbedingung kann in folgender Form geschrieben werden: 327° = y:(1— 9), wo [AuRh,] + [Au Rh, ] [Au] TRIO 3 | OP rine NER ee [AuRh, ] _ [Au Rhy] [AuRh,] + [Au Rh, } [Au] Nach dieser Bedingung ist das Verhaltnis zwischen Auro- und Aurirhodanid in den Gleichgewichtslösungen durch das Verhältnis zwischen der gesammten Goldkonzentration und der Rhodanionenkonzentra- tion bestimmt. In Fig. 2 ist diese Abhängigkeit durch eine Kurve graphisch dargestellt. In goldarmen, rhodanionenreichen Lösungen liegt das Gleichgewicht stark nach der Seite des Auro- rhodanids verschoben, und in goldreichen, rhodanionenarmen Lösungen liegt das Gleichge- wicht umgekehrt stark nach der Seite des Aurirhodanids ver- schoben. Über die Geschwindig- keit, mit welcher Aurorhodanid sich unterAusscheidung von Gold und Bildung von Aurirhodanid umsetzt, haben wir viele Mes- sungen durch Wagen des aus- #2 yg LA +1 2 7 2 geschiedenen Goldes angestellt. Es ist uns indessen nicht gelun- gen, fiir diesen Vorgang quanti- tative Zeitgesetze aufzustellen, da der Vorgang recht unregelmässig verläuft. Die Ursache hierzu ist sicher darin zu suchen, dass der Vorgang heterogen ist; heterogene Reaktionen verlaufen gewöhnlich unregelmässig, indem sie von der Oberflächenbeschaffenheit in schwierig definierbarer Fig. 2. Die Zusammensetzung der mit Gold gesättigten Goldrhodanidlösungen. 15 Weise abhangen. Wir haben gefunden, dass Goldpulver eine deutliche katalytische Wirkung auf die Goldausscheidung ausiibt. Die Geschwindigkeit der Reaktion ist so klein, dass selbst bei 40° im Laufe von 24 Stunden nur ein geringer Bruchteil des Goldes ausgeschieden wird. Der ausgeschiedene Bruchteil ist ziemlich unabhängig von der Gold- konzentration, wächst stark mit sinkender Rhodanionenkonzentration und scheint, wenigstens bei den kleineren Wasserstoffionenkonzentrationen, auch mit sinkender Wasser- stoffionenkonzentration etwas zu wachsen. Über die Deutung der Eigenschaften der Goldrhodanide mit Hilfe der gewonnenen Kenntnisse. Zu Anfang unserer Untersuchungen schienen uns die chemischen Verhält- nisse der Goldrhodanide immer komplizierter zu werden, je mehr sie untersucht wur- den; aber nach Auffindung der Gesetze der Autoreduktion und der Goldausscheidung hat es sich gezeigt, dass man durch das Zusammenspiel dieser beiden Reaktionen alle Beobachtungen erklären kaun. Durch die an und für sich reversible Goldausscheidung wird nach genügend langer Zeit jede Goldrhodanidlösung vollständig zersetzt; denn das bei der reversiblen Goldaus- scheidung gebildete Aurirhodanid wird immer aufs neue durch die irreversible Auto- reduktion entfernt. Wenn man von einer reinen Aurirhodanidlösung ausgeht, setzt die Autoreduk- tion sogleich ein; hierdurch nimmt die Konzentration des Aurirhodanids ab und die des Aurorhodanids zu. Nach einiger Zeit wird dasjenige Verhältnis zwischen Aurirhodanid und Aurorhodanid erreicht, bei dem die betreffende Lösung mit Gold in Gleichgewicht ist. Da die Autoreduktion immer weiter fortschreitet, setzt von diesem Zeitpunkt an eine Goldausscheidung ein. Diese wird nach und nach stärker, und zuletzt wird ein Zustand erreicht, bei dem die Menge des bei der Goldausscheidung gebilde- ten Aurirhodanids gerade durch das bei der Autoreduktion ver- brauchte Aurirhodanid kompensiert wird. Von jetzt an nimmt sowohl die Aurorhodanidmenge wie die Aurirhodanidmenge ständig ab. Wenn man umgekehrt von einer reinen Aurorhodanidlö- sung ausgeht, nimmt die’Aurorho- ef danidmenge ständig ab. aie die Fig. 3. Schematische Darstellung der Zustandsanderungen 2 a FR einer Aurirhodanidlösung. Aurirhodanidmenge wächst zuerst durch die reversible Goldausscheidung, bis ihr Zuwachs durch diese Reaktion von der Autoreduktion gerade kompensiert wird; in diesem Augenblick geht die Konzentration des Aurirhodanids durch ein Maximum, um danach ständig abzunehmen. In den Figuren 3 und 4 ist schematisch gezeigt, wie die Mengen von Aurirhodanid und Aurorhodanid sich mit der Zeit ändern, je nach dem man von Aurirhodanid und von Aurorhodanid ausgeht. Die von den verschiedenen Arealen abgeschnittenen Ordinaten- stücke geben die prozentischen Mengen der betreffenden Goldarten an. Sei es dass man von Aurirhodanid oder von Aurorhodanid ausgeht, so hat man nach einiger Zeit ungefähr Aurirhodand 16 dasselbe Verhältnis zwischen Auro- und Aurirhodanid; die späteren Zeiten in dem Leben der Goldrhodanidlésungen spielen sich, wie das Verhältnis zwischen Auri- und Aurogold ursprünglich auch war, in derselben Weise ab. Bei der Autoreduktion findet eine geringe Bildung von Cyanwasserstoff statt, und dieser Cyanwasserstoff wird einen kleinen Teil des Aurorhodanid- komplexes in den weit bestandigeren Aurocyanidkomplex umwandeln. Die hierdurch verursachte Kom- plikation haben wir bei der oben- gegebenen Übersicht nicht be- rücksichtigt, und ihre Berücksich- tigung Ändert die Verhältnisse auch nur unwesentlich. Fig. 4. Schematische Darstellung der Zustandsänderungen Erklärung einiger Anomalien. einer Aurorhodanidlösung. Wir schliessen diese Übersicht mit der Angabe einer Reihe von Fällen, in welchen es durch unsere Untersuchungen möglich wurde, Anomalien zu erklären, die von CLEVE oder von uns beobachtet worden sind. CLEVE fand, dass es nicht möglich ist, die einfachen Goldrhodanide darzustel- len. Wenn er Kaliumrhodanidlösung mit Aurichlorid in Überschuss oder auch nur in äquivalenter Menge versetzte, so trat immer Zersetzung ein. Nur mit Rhodanid in Über- schuss wurde diese Zersetzung verhindert, dann wurden aber nicht die einfachen Salze, sondern die Doppelrhodanide gebildet. Ebenso wenig gelang es CLEVE, das Aurorhodanid zu gewinnen. Diese Beobachtungen sind leicht zu verstehen; denn sowohl die Autoreduk- tion wie die Goldausscheidung geht bei abnehmender Rhodanionenkonzentration mit stark beschleunigter Geschwindigkeit vor sich. Für die Darstellung der festen Doppelsalze mit Aurirhodanid als Kompo- nenten ist es von Bedeutung, sich zu erinnern, dass nicht nur Rhodanionen, sondern auch Wasserstoffionen die Autoreduktion hemmen. Man soll deshalb bei der Darstellung dieser Verbindungen immer in saurer Lösung mit Überschuss von Rhodanid arbeiten. Schon CLEVE bemerkte die Wichtigkeit der Anwesenheit von Rhodanid in Überschuss; er erkannte aber nicht die Vorteile eines Arbeitens in saurer Lösung; denn bei der Dar- stellung von Kaliumaurirhodanid durch Zusatz von Aurichlorid zu Kaliumrhodanid neutralisiert er zuerst das Aurichlorid mit Kalilauge. Hierdurch verstärkt er die Auto- reduktion und vermindert die Ausbeute. Wenn man Kaliumcyanid zu einer Aurirhodanidlösung setzt, wird die Autoreduk- tion stark beschleunigt, und die Lösung verliert schnell ihre rote Farbe. Die Ursache dieser Beschleunigung ist ohne Zweifel, dass der Aurorhodanidkomplex, Au Rhy, in An- wesenheit von Kaliumeyanid in hohem Grade in Aurocyanidkomplex, Au[CN];, um- gewandelt wird, und da die Anwesenheit von Aurorhodanid auf die Autoreduktion hem- mend wirkt, muss seine Entfernung die Autoreduktion verstärken. Man könnte meinen, dass das Kaliumeyanid auch das Aurirhodanid in Cyanid umwandeln würde; eine solche Doppeldekomposition würde aber, wenn sie überhaupt stattfände, sicher momentan ver- 17 laufen, und da die rote Farbe einer Aurirhodanidlésung von Kaliumcyanid nicht augen- blicklich abgeschwächt wird, glauben wir es berechtigt, von dieser Möglichkeit abzusehen. CLEVE fand, dass eine alkoholische Lösung von Kaliumaurirhodanid mit Silber- nitrat einen Niederschlag gibt, der zuerst rotbraun und gelatinös ist, aber schnell weiss und käsig wird, und in dieser Form annähernd die Zusammensetzung Ag,Au Rh, be- sitzt. CLEvE deutet an, dass der Niederschlag als eine Mischung von Silberaurorhodanid und Silberrhodanid, AgAu Rh, + 2 Ag Rh, aufzufassen ist. Die Bildung einer ähnlichen Mischung ist nach unseren Untersuchungen auch zu erwarten. Durch Zusatz von Silber- nitrat entfernt man aus der Lösung sowohl die Rhodanionen wie die Dirhodanoauro- ationen als Silbersalze, und man muss somit eine sehr schnelle Autoreduktion erwarten. Daher wandelt sich der ursprüngliche Niederschlag von Silberaurirhodanid schnell in eine Mischung von Silberaurorhodanid, Silberrhodanid und Silbercyanid um: 3AgAuRh, +6AgN0, + H,0 — 3AgAuRh, +5AgRh+AgCN + H,SO,. Nur noch eine Beobachtung soll erwähnt werden. Wenn eine konzentrierte, saure Lösung von Kaliumaurorhodanid aufbewahrt wird, scheiden sich in einigen Tagen schön tiefrote Krystalle aus, die annähernd die Zusammensetzung KAu,Rh, besitzen. Wir meinten deshalb zuerst, ein neues Kaliumaurorhodanid gefunden zu haben. Die nähere Untersuchung der Krystalle zeigte aber, dass es trotz des schönen und homogenen Aussehens nur eine Mischung von rotem Kaliumaurirhodanid und Gold, KAuRh, + 2Au, war; denn durch Pulverisieren und Schlemmen mit etwas Wasser konnten wir schweres metallisches Goldpulver von den leichteren Kaliumaurirhodanidpartikeln trennen, und durch Behandlung mit Quecksilber konnten die Krystalle fast quantitativ in Goldamalgam und Kaliumaurirhodanid umgewandelt werden. Das bei der Zersetzung der Kalium- aurorhodanidlösung entstandene Gold und Kaliumaurirhodanid wird also wegen der grossen Schwerlöslichkeit des Kaliumaurirhodanids in der Form von »Pseudomisch- krystallen« ausgeschieden. Über die Anwendung der Goldrhodanide zu Tonbädern. Wir haben bei unseren Unter- suchungen auf die Anwendung der Goldrhodanide in der photographischen Technik kein besonderes Gewicht gelegt. Die folgenden Bemerkungen sind deshalb nicht als Re- sultate besonderer Untersuchungen aufzufassen, sondern enthalten nur theoretische Er- klärungen einiger aus der Technik bekannten Erscheinungen. Wenn man durch Zusatz einer geringen Menge Aurichlorid zu einer verdünnten Lösung von Ammoniumrhodanid ein Goldtonbad herstellt, arbeitet man unter Bedin- gungen (wenig Wasserstoff- und Rhodanionen), bei welchen die Autoreduktion sehr schnell verläuft, und man sieht denn auch, dass die Lösung schnell ihre rote Farbe verliert. Es wird immer in den Rezepten für goldrhodanidhaltige Tonbäder bemerkt, dass die Lösung nicht frisch angewendet werden darf. Die Zweckmässigkeit dieser Forderung ist leicht zu verstehen; denn da die Rhodanidkonzentration in den Lösungen immer weit grösser ist als die Goldkonzentration, muss die Aurirhodanidkonzentration auf einen sehr kleinen Wert sinken, ehe die Lösung den Punkt erreicht, wo sie mit metallischem Gold in Gleich- gewicht ist. Und es ist natürlich für ihre Anwendung als Tonbad wichtig, dass dieser Punkt erreicht oder vielmehr überschritten wird. Bevor dieser Punkt erreicht ist, wirkt die Lösung nämlich goldauflösend und schwächt also das photographische Bild, wäh- D.K.D. Vidensk Selsk. Skr., naturvidensk. og mathem. Afd., 8, Række, V.1, 3 18 rend sie nach dem Uberschreiten dieses Punktes goldausscheidende Eigenschaften erhält und also verstärkend wirken kann. Allzu lange darf man aber auch nicht ein Goldrho- danidtonbad aufbewahren, da dann das anwesende Gold als Metall ausgeschieden wird. EDER empfiehlt, dem Goldrhodanidtonbad einen Zusatz von Natriumazetat zu geben. Es scheint uns nicht unwahrscheinlich, dass dieser Zusatz durch die von ihm hervor- gerufene Verkleinerung der Wasserstoffionenkonzentration und die damit verbundene grössere Geschwindigkeit der Goldausscheidung die verstärkenden Eigenschaften des Tonbads vergrössern wird. Wenn Goldrhodanid im Vergleich mit vielen anderen Goldsalzen, z. B. Goldchlorid, als Tonbad besonders geeignet ist, so ist es dabei sicher von Bedeutung, dass das Gold darin einwertig vorhanden ist. Aus einer Goldlösung mit einwertigem Gold scheidet eine bestimmte Silbermenge dreimal so viel Gold aus als aus einer Lösung, in welcher das Gold dreiwertig ist. ll. Darstellung und Eigenschaften der in fester Form gewonnenen goldrhodanidhaltigen Verbindungen. 1. Die krystallisierten Tetrarhodanoauriate. Kaliumaurirhodanid, KAuRh, (Kaliumtetrarhodanoauriat) haben wir mehr- mals nach CLEVE dargestellt. Das ausgeschiedene Salz wurde mit Wasser gewaschen und über Schwefelsäure getrocknet. Es wurden drei verschiedene Präparate analysiert, und in allen war der Goldgehalt ein wenig zu gross und der Kaliumgehalt etwas zu klein. Berechnet nach KAuRh, I II Ill °/o Au 42,08 42,43; 42,44; 42,38 42,40 42,36; 42,27 mi 330 8,34 8,06; 8,02 8,16 - S 27,37 27,63 Gold und Kalium wurden nach Abrauchen mit Schwefelsåure als Gold + Kalium- sulfat gewogen, und nachdem das Kaliumsulfat mit Wasser entfernt war, wurde das Gold für sich gewogen. Schwefel wurde nach Oxydation mit Kônigswasser als Baryum- sulfat gefallt und gewogen. Kaliumaurirhodanid ist sehr schwerlöslich in Wasser. Eine Mischung von der Zu- sammensetzung: 1m HCl, 0,4m KRh, 0,005m HAuCl, (reduziert zu Aurogold mit Sul- fit), 0,001 m HAuCI,, also eine Mischung mit 0,005m Aurorhodanid und 0,001 m Auri- rhodanid, schied sogleich das Kaliumsalz aus, und die nur schwach gefärbte Lö- sung wies bei Sulfittitrierung nach Schütteln und Absetzen in drei Minuten bei ca. 22° einen Inhalt von 0,0002 m Aurigold auf. Hieraus berechnet sich das Löslichkeitspro- dukt des Salzes zu 0,00008. [K*]-[AuRh,] = 0,4 + 0,0002 = 0,00008. Eine Mischung von 1m HCl, 0,4m K Rh, 0,001 m HAuCl,, welche von der vorher- gehenden nur durch das Fehlen von Aurogold abweicht, schied auch das Kaliumsalz KJ 19 aus und wies nach Schiitteln und Stehen in drei Minuten bei 22° einen Inhalt von 0,0003 m Aurigold auf. Dieses grössere Titrierungsresultat findet seine Erklärung durch die Disso- ziation des Aurirhodanids in Aurorhodanid und freies Rhodan. Der Dissoziationsgrad, a, des Aurirhodanids in der letzten Lösung berechnet sich nach der Dissoziationsgleichung (S. 10) zu 0,33: a [AuwRh, ]- [Rh] — 049.107" — : See [Au Rk] — 049 10° = 70,0008; a — 0,33. Von dem durch die Sulfittitrierung angezeigten 0,0003 m Aurirhodanid ist also in Wirklichkeit nur 0,0002 m als Aurirhodanid vorhanden, der Rest, 0,0001 m, ist als freies Rhodan anwesend. Sachgemäss berechnet, ergibt daher auch dieser Versuch für das Lés- lichkeitsprodukt des Kaliumaurirhodanids den Wert 0,00008. In der ersten Lösung ist die Dissoziation von dem vorhandenen Aurorhodanid so stark zurückgedrängt, dass man sie nicht zu berücksichtigen braucht. Der Wert 0,00008 für das Löslichkeitsprodukt gilt für Lösungen, in welchen die Ak- tivitätskoeffizienten der Ionen dieselben sind wie in den hier untersuchten Lösungen, d. h. für Lösungen mit einer Ionenkonzentration von etwa 1,4 m. Für andere Lösungen muss man die Änderungen der Aktivitätskoeffizienten berücksichtigen, wenn man den genauen Wert der Löslichkeit zu berechnen wünscht. Ammoniumaurirhodanid, NH,AuRh,, (Ammoniumtetrarhodanoauriat) wurde durch Fällen einer Lösung von 3 g Ammoniumrhodanid in 250 cm? Wasser mit 250 cm? einer salzsauren Lösung von Aurichlorid (mit 0,4 °/o Gold) gewonnen. Die Flüssigkeit wurde sogleich rotbraun und schied fast augenblicklich einen rotgelben, voluminösen Niederschlag aus, welcher dem Kaliumaurirhodanid zum Verwechseln ähnlich war; der Niederschlag wurde abgesaugt, zweimal mit 0,01 m Ammoniumrhodanid und zweimal mit Wasser gewaschen und über Schwefelsäure getrocknet (Präparat I). Die Ausbeute war 1,04 g (46 io). Eine andere Portion des Salzes wurde in konzentrierterer Lösung gefällt (Präparat II). Die beiden Produkte besassen genau die Formel NH,AuRh,. Berechnet nach NH, Au Rh, Kos II °/o Au 44,06 44,02; 44,05 44,14 Ss 28,65 28,7; 28,42 28,75 ;28,23 ;28,74. Natriumaurirhodanid, NaAuRh,, (Natriumtetrarhodanoauriat) wurde dar- gestellt durch Fällen einer verdünnten Lösung von Natriumrhodanid (ca. 0,1 m) mit 0,1 molarem Wasserstoflaurichlorid in zur vollständigen Fällung ungenügender Menge. Hierdurch wurde es als ein roter, schön krystallinischer Niederschlag erhalten, der ab- gesaugt und über Schwefelsäure getrocknet wurde. Berechnet nach NaAuRh, Gefunden °/o Au 43,58 43,1; 43,5 = S 28,34 28,18; 28,3 Aus einer schwach übersättigten salzsauren Lösung scheidet sich dieses Salz lang- sam in schönen, millimetergrossen, rubinroten Nadeln aus. 3* 20 Die Léslichkeit des Natriumaurirhodanids kann nach folgenden Versuchen geschatzt werden. Eine Lösung, 0,5m NaRh, 1m HCl, 0,004m HAuCl,, schied bei gewöhnlicher Temperatur nach einer Minute Krystalle von Natriumaurirhodanid aus, während eine entsprechende Lösung mit nur 0,4m NaRh sich klar hielt. Eine Lösung !/„m H AuCl,, 1/,m NaRh, 1/,m HCl schied nach 45 Sekunden Krystalle aus. Nach diesen Angaben muss das Löslichkeitsprodukt des Natriumaurirhodanids wahrscheinlich etwas kleiner als 0,002 sein. [Nat] -[AuRhz] = 0,5 : 0,004 = 0,002; [Nat] [AuRhz] = !/, % = 0,002. Das Löslichkeitsprodukt des Natriumsalzes ist also ca. 20mal grösser als das des Ka- liumsalzes. Das Natriumaurirhodanid soll von Kern!) durch Fällen einer konzentrierten Lösung von Natriumgoldchlorid mit einer Lösung von Ammoniumrhodanid dargestellt worden sein. Da das Natriumsalz indessen weit löslicher ist als Ammoniumaurirhodanid, muss man nach dem Verfahren von Kern einen Niederschlag von Ammoniumaurirhodanid erwarten. Es war uns nicht möglich, die Originalabhandlung von Kern zu erhalten, und wir haben deshalb nicht versucht, seine Darstellung zu wiederholen. Wasserstoffaurirhodanid, HAuRh, 2H,0, (Tetrarhodanoauriatsäure) wurde in folgender Weise erhalten. 30 g Natriumrhodanid (NaRh, 2H,0) wurden in einem Schei- detrichter in 200 cm? verdünnter Schwefelsäure (ca. 1,5 m) gelöst und mit 100 cm? Wasser- stoffaurichloridlösung (10 °/o Gold) vermischt. Hierbei schied sich Natriumaurirhodanid in grosser Menge aus. Die ganze Mischung wurde mit 150 cm? Äther geschüttelt. Hier- durch bildeten sich eine dunkelrote ätherische Schicht und eine hellrote wässrige Schicht. Die ätherische Schicht wurde abgetrennt, mit 15 g entwässertem Natriumsulfat 10 Mi- nuten getrocknet und im Vakuum bei ca. 20° eingedampft. Hierbei schied sich die Säure in schönen, dunkelroten Krystallen aus. Die ausgeschiedenen Krystalle wurden abgetrennt, auf Goochtiegel abgesaugt, durch Zentrifugieren so vollständig wie möglich von Mutterlauge befreit und in der Luft getrocknet. Es wurden im ganzen vier nach- einander aus der ätherischen Lösung ausgeschiedene Produkte gewonnen und analysiert. °/o Au oS Berechnet nach HAuBh, 32 BO) Ir ee 42,26 27,48 Gefunden. im ersten Produkt...) sr ee 42,16 Wy pee byt =e - zweiten (==: VU). Re. ate eee eee tee 42,15 27,61 — - dritten -— | LT AAR NE KFS RENEE SER 41,94 — - v vierten’. — 1 UNE fees. ARENSE cia 42,0 Gold und Schwefel wurden nach Oxydation mit Königswasser durch Fällen mit Schwefeldioxyd, bzw. Baryumchlorid bestimmt. Die Zusammensetzung des ersten Produktes stimmt genau mit der Formel H Au Rh,, 2H,0. Beim Liegen in freier Luft ändert die Säure ihr Gewicht nicht. Dagegen verliert sie schnell an Gewicht über Schwefelsäure. Hierbei gehen nicht nur die zwei Moleküle 1) Chem. News 83, 243 (1876). Ref.: Ber. deut. chem. Ges. 8, 1684 (1875). Jahresberichte 1876, 319. 21 Wasser weg, sondern es tritt eine tiefergehende Zersetzung ein. In vier Tagen wurden 10,0°/) an Gewicht abgegeben; den zwei Molekülen Wasser würden nur 7,7 °/, ent- sprechen. 2. Die krystallisierten Dirhodanoauroate und das Monamminoaurorhodanid. CLEVE hat aus den Zersetzungsprodukten des Kaliumaurirhodanids in heisser, wäss- riger Lösung ein Kaliumaurorhodanid isoliert. Er empfiehlt zur Darstellung dieses Dop- pelsalzes eine neutralisierte Lösung von Aurichlorid zu einer 80° warmen Lösung von Kaliumrhodanid zu setzen; jeder Zusatz von Aurichlorid ergibt einen Niederschlag von Kaliumaurirhodanid, und erst wenn dieser Niederschlag gelöst ist, wird mehr zugesetzt; zuletzt wird die gelbe Lösung zur Krystallisation eingedampft, wodurch eine sehr unreine Krystallmasse ausgeschieden wird; aus dieser Masse wurde, wenn auch nur schwierig, ein strohgelbes, krystallinisches, in Wasser und Alkohol leichtlösliches Salz isoliert, dessen Analyse mit der Formel KAuRh, stimmte. Wir haben versucht, die CLeve’sche Prä- paration zu wiederholen; es ist uns aber nicht gelungen, eine auch nur annähernd reine Verbindung zu erhalten, und da das Kaliumaurorhodanid, welches wir in anderer Weise rein dargestellt haben, ein rein weisses Aussehen besitzt, glauben wir, dass das strohgelbe Produkt von CLEVE auch ziemlich unrein war. Statt wie CLEVE das Aurirhodanid durch Autoreduktion in heisser Lösung in Auro- rhodanid umzuwandeln, kann man es bei gewöhnlicher Temperatur mit Sulfit reduzieren; dadurch vermeidet man die Bildung unangenehmer Nebenprodukte und erhält eine farblose Lösung von Aurorhodanid. Um daraus feste Aurorhodanide zu gewinnen, extrahiert man die reduzierte, angesäuerte Lösung mit Äther und erhält dabei die Dirhodanoauroatsäure in ätherischer Lösung, und durch geeignete Neutralisation dieser Lösung gewinnt man reine, krystallinische Dirhodanoauroate. In dieser Weise haben wir durch Neutralisation mit Kaliumbikarbonat das Kaliumdirhodanoauroat, KAuRh,, und durch Neutra- lisation mit Ammoniak das Ammoniumdirhodanoauroat, NH,AuRh,, gewonnen. Beide Salze sind weiss, krystallinisch umd leicht löslich in Wasser und Alkohol. Wenn man mit der Zuleitung von Ammoniak fortsetzt, nachdem die ätherische Lösung von der Dirhodanoauroatsäure neutralisiert ist, scheidet sich nach und nach eine weisse, schwer- lösliche Verbindung aus, nämlich das Monamminoaurorhodanid, AuNH,Rh: NH,AuRh, + NH, = AuNH,Rh + NH,Rh. Diese Verbindung ist auch schon von CLEVE dargestellt, aber nur in sehr kleiner Menge und nach einem sehr beschwerlichen Verfahren. Wenn man eine angesäuerte Aurorhodanidlösung mit Äther ausschüttelt und die ätherische Schicht mit säurehaltigem Wasser wäscht, bis das Waschwasser nur eine schwache Rhodanreaktion mit Ferrisalz ergibt, enthalt die Atherschicht zwei Rhodan- gruppen pro Goldatom, und man darf daher annehmen, dass sie die Dirhodanoauroat- säure, H AuRh,, in reinem Zustande enthält. Durch Eindampfen der ätherischen Schicht gewinnt man ein rötliches Öl. Die rötliche Farbe, die mit der Zeit zunimmt, stammt sicher von einem Inhalt von Aurirhodanid. In Äther geht das Öl teilweise in Lösung und lässt einem weissen, amorphen Niederschlag zurück, den wir nicht näher untersucht 22 haben. Die Zersetzung kann doch keine tiefgreifende sein; denn wenn man Lésung plus Niederschlag mit Ammoniak neutralisiert, geht alles in Lésung, und durch Ein- dampfen zur Krystallisation gewinnt man schöne, weisse, analysenreine Krystalle von Ammoniumdirhodanoauroat. Nach eintägigem Trocknen über Phosphorpentoxyd wur- den 59,4 °/o Gold gefunden, während die Formel NH,AuRh, 59,5 °%/o Gold verlangt. Kaliumaurorhodanid, KAuRh, (Kaliumdirhodanoauroat). 6 g Kaliumrho- danid wurden in einem Scheidetrichter in 150 cm? Wasser gelöst und 50 cm? 0,5 molares Wasserstoffaurichlorid in kleinen Portionen zugesetzt; jede Portion wurde mit einer konzentrierten Natriumsulfitlösung entfärbt, ehe die nächste zugesetzt wurde. Nach Zusatz von allem Wasserstoflaurichlorid wurde die Lösung mit etwas verdünnter Schwe- felsäure versetzt und mit Äther ausgeschüttelt. Die ätherische Schicht wurde einigemal mit stark verdünnter Schwefelsäure gewaschen, bis das Waschwasser mit Ferrisalz nur schwache Rhodanreaktion ergab, und danach mit festem Kaliumbikarbonat (Kalium- hydroxyd wirkte zersetzend) in Überschuss versetzt; hierdurch wurde Kohlendioxyd entwickelt und ein voluminöser, aus einer Mischung von Kaliumaurorhodanid und Kalium- bikarbonat bestehender Niederschlag gebildet. Nach Abgiessen des Äthers wurde das Ka- liumaurorhodanid aus dem Niederschlag mit ein wenig absolutem Alkohol extrahiert, und die erhaltene Lösung wurde im Vakuum über Schwefelsäure eingedampft. Nach kurzer Zeit schied sich das Salz in feinen, weissen Nadeln aus. Es wurde abfiltriert, mit wasserfreiem Äther gewaschen und über Phosphorpentoxyd getrocknet. Das erhaltene Salz war ein weisser, krystallinischer Körper, sehr leicht löslich in Al- kohol, schwer löslich in Äther. In Wasser wurde es unter anfangender Zersetzung gelöst. Bei der Analyse gab es folgende Resultate: Berechnet nach Gefunden ') KAuRh, I II Ill %o Au 59,95 56,00 56,01 55,88 "NS 18,2 18,1 18,49 - K ane 112 Ammoniumaurorhodanid, NH,AuRh,, (Ammoniumdirhodanoauroat) wurde in ähnlicher Weise wie das entsprechende Kaliumsalz dargestellt; zuerst wurde eine athe- rische Lésung von Wasserstoffaurorhodanid dargestellt und diese dann genau mit Am- moniakgas neutralisiert. Wenn man in der Lösung ein wenig Aurigold zurücklässt, so dass sie eine schwache rotbraune Farbe besitzt, kann man mit der Zuleitung von Ammoniak aufhören, wenn diese Farbe verschwindet, denn da Aurirhodanid nur in saurer Lösung haltbar ist, ist das Verschwinden seiner Farbe ein Zeichen der eingetretenen Neutrali- sation. Nach viertelstündigem Trocknen über entwässertem Natriumsulfat wurde die Lösung zur Krystallisation eingedampft. Die ausgeschiedenen Krystalle wurden abfiltriert, mit Äther gewaschen und über Phosphorpentoxyd getrocknet. Ausbeute 7,1 g, berech- net nach der angewandten Goldmenge 8,4 g. Das erhaltene Salz war schön weiss und krystallinisch. Es ist frisch dargestellt leicht löslich in Äther, aber nach Trocknen sehr schwer löslich darin. 25 g wasserfreies Äther 1) Ausführung der Analysen, siehe Seite 18. 23 enthielt nach 24stiindigem Schütteln mit 0,5 g getrocknetem Salz bei 25° nur 0,013 g Gold. In Alkohol ist es leicht löslich, doch mit einem rötlichen Farbton, der eine geringe Bildung von Aurirhodanid andeutet. In Wasser ist es auch leicht löslich, die Lösung scheidet aber nach einiger Zeit einen gelben Niederschlag aus, was eine Zersetzung an- deutet. Die Analyse ergab folgende Resultate. I og II sind mehrere Tage über Phosphor- pentoxyd getrocknet, III ist aus wasserhaltigem Äther umkrystallisiert und getrocknet, IV ist aus Alkohol umkrystallisiert und getrocknet. Berechnet nach Gefunden NH,AuRh, I Il III IV °/o Au 59,50 59,58 59,5 59,46 59,56 = 19,33 19,5 19,6 - NH, 5,14 5,13 5,10 Die Bestimmungen von Gold und Schwefel wurden wie gewöhnlich ausgeführt (siehe Seite 18). Ammoniak wurde nach Zusatz von Natronlauge abdestilliert, in Säure auf- gefangen und durch Zurücktitrierung mit Natriumhydroxyd bestimmt. Vor dem Zu- satze der Natronlauge ist es notwendig, das Salz mit Schwefeldioxydwasser zu redu- zieren; sonst oxydiert das Aurogold in der alkalischen Lösung etwas Ammoniak, und man findet statt über 5 °/o kaum 4 /o. Monamminoaurorhodanid, AuNH,Rh. Wenn man die ätherische Lösung von Dirhodanoauroatsäure nach Entwässerung mit wasserfreiem Natriumsulfat mit Ammoniak sättigt, erhält man einen weissen, chlorsilberähnlichen Niederschlag. Nach Abfiltrieren, Waschen mit Alkohol und Äther und Trocknen im Vakuum über Schwefel- säure wurde der Niederschlag analysiert. Berechnet nach Gefunden AuNH,Rh I II Ill % Au 72,42 72,53 72,44 72,42 = S 11,78 11,67 11,86 11,84 - NH, 6,25 6,22 Monamminoaurorhodanid ist recht unbeständig. Es wird im Lichte geschwärzt und von Wasser zersetzt. Wenn man bei seiner Darstellung die ätherische Lösung der Dirhodanoauroatsäure nicht trocknet, schwärzt sich die gefallte Amminoverbindung etwas, ehe sie getrocknet werden kann. Die Verbindung ist sehr schwerlôslich in wasser- freiem Äther und wasserfreiem Alkohol, aber etwas löslich in gewöhnlichem Alkohol. In ammoniumrhodanidhaltigem Alkohol ist die Verbindung leicht löslich unter Bildung von Ammoniumaurorhodanid und Ammoniak; wenn man daher das mit Ammoniak gefällte Monamminoaurorhodanid mit Alkohol wäscht, geht anfangs ein Teil in Lösung; denn der Niederschlag enthält Ammoniumrhodanid, das mit Alkohol ausgewaschen wird und dabei lösend wirkt. Wenn man mit ammoniakhaltigem Alkohol wäscht, wird man wahrscheinlich diesen Verlust vermeiden können. 24 Ill. Zusammensetzung und Komplexität des Aurorhodanidkomplexes. Potentialmessungen nach BODLÄNDER in Aurorhodanidlösungen, Die Zusammen- setzung der festen Rhodanoauroate macht es wahrscheinlich, dass das Gold in rho- danidhaltigen Aurolösungen als Dirhodanoauroation, AuRh,, vorhanden sein muss. Um indessen über die Zusammensetzung des in den Lösungen vorhandenen Rhodano- auroatkomplexes Sicherheit zu erhalten, haben wir seine Zusammensetzung nach einem Verfahren von BopLÄnDER!) bestimmt und haben dadurch bestätigt gefunden, dass der Komplex ein Dirhodanoauroation, Au KRh,, ist. Bei dem Verfahren von BopLANDER misst man das Potential von Konzentrations- ketten unter Anwendung des in dem Komplex enthaltenen Metalls als Elektroden. Man macht einen Versuch, in welchem der Gehalt der Lösungen an dem Komplex an den beiden Elektroden verschieden ist, während der nicht metallische Bestandteil des Kom- plexes in gleicher Menge in beiden Lösungen in Überschuss vorhanden ist, und einen anderen Versuch, in welchem die Lösungen den Komplex in gleichen Mengen, den Über- schuss der nicht metallischen Komponente aber in verschiedenen Mengen enthalten. Aus den Werten dieser Potentiale kann man die Zusammensetzung des Komplexes berechnen. Um dieses Verfahren auf den Aurorhodanidkomplex anzuwenden, muss man das Potential zwischen zwei Goldelektroden messen, die in zwei verschiedene Aurorho- danidlösungen eintauchen. Wenn man in diesen zwei Lösungen dieselbe Rhodanionen- konzentration und eine verschiedene Aurogoldkonzentration anwendet, erhält man aus der Potentialmessung die Zahl der Goldatome im Komplex, und wenn man dieselbe Au- rogoldkonzentration und eine verschiedene Rhodanionenkonzentration anwendet, er- hält man aus der Potentialmessung die Zahl der Rhodangruppen pro Goldatom im Komplex. Die Tabellen 4-—6 enthalten die Resultate unserer Versuche. Die Konzentration des Goldes ist überall in dezimillimolaren Konzentrationen (dmm) angegeben. E ist das gemessene Potential in Volt, n und m die Zahl der Rhodanradikale, bzw. der Goldatome im Komplex (AumRhn). Die angeführten Werte von n:m und m sind nach folgenden Formeln berechnet: n E 1,983 - 10“. T de mn. 11088 107 Ta CU Ei, ob In der ersten Formel bedeuten C, und C, die Rhodanionenkonzentrationen in Elek- trodenlösungen mit derselben Goldkonzentration, und in der zweiten Formel sind €, und C, die Goldkonzentrationen in Elektrodenlésungen mit derselben Rhodanionen- konzentration. Das Mittel aus den Werten der Tabellen für m und n; m ist: m = 1,02. und n2m = 1,9% Dies bedeutet, dass der Rhodanoauroatkomplex in den untersuchten Lösungen die For- mel AuRhz besitzt. In den untersuchten Lösungen variiert die Rhodanionenkonzen- *) DEDERInD-Festschrift, 162 (1901). BopLÄnper u. FirriG: Zeitschr. physik. Chem. 39, 607 (1902). BoDLÄNDER u. StonBEck: Zeitschr. f. anorg. Chem. 31, 1, 458 (1902). 25 tration von 0,1 bis 1 molar und die Aurogoldkonzentration von 1 bis 40 dezimillimolar. Auf diesem grossen Gebiet hat also der Rhodanoauroatkomplex die angegebene Zusammen- setzung. In den untersuchten Lösungen variiert die Säuremenge auch bedeutend (von etwa 0,0003 bis 1 molar). Tabelle 4. Potential von Ketten : Au | Lösung I | gesätt. KC1| Lösung II | Au. Die Lösungen enthalten das Gold als durch Autoreduktion gebildetes Aurogold. Lösung I Lösung II Temp. E nlm 1 dmm Au, 0,1 m KRh 1 dmm Au, 1 m KRh 19,5° 0,1212 2,09 1 — 0,1 — 1 — 0,5, — — 0,0875 2,16 1 _ 0,5 — 1 — 1 — — 0,0357 2,05 2,5 — 0,25 — 25 — 0,5 — — 0,0345 1,98 5 — 0,5 -- 5 — 1 — — 0,0375 2,15 m 2,5 dmm Au, 0,5 m KRh 0,625 dmm Au, 0,5 m KRh 202 0,0304 » 1,15 5 — 0,5 — 1,25 — 05 — — 0,0308 1,13 BE Ke a 1 ih Hyannis = 0,0460 0,88 ARE Ls 0,625 — 1 = = 0,0357 0,98 5 — 1 — 1,25 — 1 = — 0,0373 0,94 10 — 1 — 2,5 — 1 — — 0,0369 0,95 Dines et = 1 très = 19,5° 0,0242 0,95 5 — 1 — 1 — 1 — — 0,0421 0,96 Bo er 2,5 = 1 == = 0,0173 1,01 10 — 1 — 1 — 1 — — 0,0588 0,99 Tabelle 5. Potential von Ketten : Au | Lösung I | gesått. KC/ | Lösung II | Au. Die Lösungen enthalten das Gold als durch Autoreduktion gebildetes Aurogold. Lösung I Lösung II Temp. E Nim 8,35 dmm Au, 0,25 m KRh 8,35 dmm Au, 0,5 m KRh 23° 0,0338 1,91 8,35 — 0,5 — 8,35 — 1 — — 0,0329 1,86 8,35 — 0,25 — 8,35 = 1 = == 0,0663 1,88 16,7 — 0,5 — 16,7 — 1 — — 0,0359 2,03 Tabelle 6. Potential von Ketten : Au | Lösung I | gesatt. KCI | Lösung II | Au. Die Lösungen enthalten das Gold als durch Reduktion mit Sulfit gebildetes Aurogold. Lösung 1 Lösung II Temp. E Nim 40 dmm Au, 0,4 m NaRh, 1m HCl 40 dmm Au, 0,1 m NaRh, 1m HCl 20° 0,0634 1,81 40 — 0,4 = 1° 400 — 0,2 — 1 — — 0,0338 1,94 m 40 dmm Au, 0,4 m NaRh, 1m HCl 20 dmm Au, 0,4 m NaRh, 1 m HCl 20° 0,0139 1,25 Die Ausführung der Versuche. Die angewandten Aurorhodanidlösungen wur- den aus Aurirhodanid dargestellt teils durch Autoreduktion und teils durch Reduktion mit Sulfit. In der ersten Messungsserie (Tabelle 4) wurde zur Darstellung der Elektro- denlösungen eine Aurorhodanidlösung A benutzt, die durch Schütteln von 300 cm? 1 molarem Kaliumrhodanid mit 0,1406 g Kaliumaurirhodanid bis zur vollständigen D. K. D. Vidensk, Selsk. Skr., naturvidensk. og mathem. Afd. 8. Række, V, 1. 4 26 Lösung gebildet war. Das schwerlösliche Kaliumaurirhodanid geht nach und nach in Lösung, indem es in der nur schwach sauren Lösung schnell zu Avrorhodanid autoredu- ziert wird. Die hierdurch gebildete 0,001 molare Aurorhodanidlösung besass eine ganz schwache rotbraune Farbe und enthielt also eine Spur von Aurirhodanid. In einer zwei- ten Serie (Tabelle 5) wurde zur Darstellung der Elektrodenlösungen eine Aurorhodanid- lösung B angewandt, die durch zweitägiges Schütteln von 300 cm? 1 molarem Kalium- rhodanid mit 0,8 g Kaliumaurirhodanid gebildet war. Hierdurch wurde nicht alles ge- löst; nach einem Tag wurde in 50 cm? 0,028 g Gold gefunden, und nach zwei Tagen 0,033 g Gold in 50 cm. Die Lösung B war hiernach 0,00334 m Au; sie war bedeutend rotbrauner als A und enthielt also bedeutend mehr Aurigold. Aus A und B wurden die Elektrodenlösungen durch Verdünnung mit Wasser oder mit Kaliumrhodanid dar- gestellt. Eine dritte Serie (Tabelle 6) umfasst Versuche, in welchen das Aurogold durch Reduktion mit Sulfit dargestellt wurde. Die hier benutzten Lösungen wurden aus 2 mo- larer Salzsäure, 1 molarem Natriumrhodanid und 0,04 molarem, mit Natriumhydroxyd neutralisiertem Goldchlorid dargestellt. Nach Mischung von passenden Mengen von die- sen Lösungen wurde mit gerade der notwendigen Menge Natriumsulfit entfärbt und mit Wasser auf das gewünschte Volumen verdünnt. Als Elektroden wurden galvanisch vergoldete, in Glasröhre eingeschmolzene Platindrähte angewandt. Die Potentiale wur- den nach der Kompensationsmethode gemessen, unter Anwendung eines 100 cm langen Messdrahts, eines Weston-Normalelements und eines Spiegelgalvanometers von SIE- MENS und HALSKE. In den Elektrodenlésungen sind die Rhodanionenkonzentrationen aus den ange- wandten Mengen Kalium- oder Natriumrhodanid berechnet, ohne Beriicksichtigung der durch die Autoreduktion von Kaliumaurirhodanid gebildeten Rhodanionen oder der zur Bildung von Rhodanokomplex aus Aurichlorid verbrauchten Rhodanionen. Eine solche Berücksichtigung würde die Konzentrationen nur unwesentlich ändern. | Weiter haben wir statt der in den Formeln enthaltenen Aurorhodanidkonzentration die Gesamtkonzentration des Goldes benutzt. In den sulfitreduzierten Lösungen ist die- ses Verfahren ganz genau, da alles Gold wirklich als Aurorhodanid vorhanden ist; in den autoreduzierten Lösungen ist aber ein Teil des Goldes als Cyanokomplex*) und ein anderer Teil des Goldes als Aurirhodanid vorhanden, und der Ansatz ist daher nicht ganz exakt; er ist aber zulässig, weil der Cyanokomplex immer ungefähr denselben Bruchteil der ganzen Goldmenge ausmacht, und weil das Aurirhodanid nur in geringer Menge vorhanden ist. Das Normalpotential Gold-Aurorhodanid. Die soeben beschriebenen Messungen haben uns gezeigt, dass das Potential einer Goldelektrode in Aurorhodanid nach der Formel [Au Rk, ] E = E+7 2 Im [Rh } mit der Zusammensetzung der Lösung variiert. Da indessen nur Messungen von Gold- Aurorhodanid-Elektroden gegeneinander ausgeführt wurden, können wir aus diesen 1) Vergleich später Seite 28, 27 Messungen „E, das Normalpotential Gold-Aurorhodanid, nicht bestimmen. Um diese Lücke auszufüllen, haben wir auch einige Messungen von Gold-Aurorhodanid-Elektro- den gegen Kalomel-Elektroden ausgefiihrt. Bei diesen Messungen wurde das Gold in Form von mit Natriumhydroxyd neutralisiertem Wasserstoffaurichlorid zugesetzt und mit Natriumsulfit genau reduziert. Tabelle 7 enthalt die Resultate unserer Messungen. Tabelle 7. Potential von Ketten : Au | Aurolôsung | gesätt. KCl) 0,1 m KCl, Hg Cl | Hg. Aurolösung Temp. [Rh] E E’ oE 0,004 m Au’, 0.4m NaRh, 1m HCl 20° 0,392 0,2423 0,2583 0,3503 0,004 m — , 02m — ‚lm — 20° 0,192 0,2787 0,2947 0,3507 0,0022m —,04m — ‚lm — 20° 0,396 0,2287 0,2447 0,3548 Mittel: 0,352 In der Tabelle sind zuerst die Zusammensetzungen und die Temperaturen der Lö- sungen angegeben. Dann folgen die Rhodanionenkonzentrationen der Lösungen, [Rh]; sie sind berechnet aus den angewandten Mengen von Natriumrhodanid, indem für jedes Goldatom zwei Rhodanradikale abgezogen worden sind. E ist das gemessene Potential in Volt, E' ist das für das Diffusionspotential korrigierte Potential; nach der HENDERSON’ schen Formel!) findet man für das Diffusionspotential zwischen 1 molarer Salzsäure und gesättigter Kaliumchloridlösung den Wert 0,016 Volt; die Anwesenheit vom Natrium- rhodanid und den anderen Bestandteilen der Lösungen kann diesen Wert nur unwe- sentlich ändern, und E’ ist daher überall aus E durch Addition von 0,016 berechnet. oE ist das aus E' berechnete Normalpotential Gold-Aurorhodanid; es ist berechnet nach der folgenden Formel: Für [AuRhz] ist die gesamte Goldkonzentration benutzt und für [Rh] die in der Tabelle angegebene Zahl. Das Mittel aller Werte des Normalpotentials ist + 0,352 Volt. Wenn man für das Potential der 0,1 normalen Kalomel-Elektroden gegenüber der Nor- mal-Wasserstofi-Elektrode den Wert +0,337 Volt benutzt, erhält man für das Normal- potential Gold-Aurorhodanid, gemessen gegen die Normal-Wasserstoff-Elektrode: oE, = + 0,689 Volt Ehe wir die Bedeutung dieses Wertes diskutieren, wünschen wir noch einige Be- merkungen zu machen, die die Zusammensetzung derjenigen Aurorhodanidlösungen betreffen, in welchen wir eine wohldefinierte Messung von dem Potential Gold-Auro- rhodanid erwarten dürfen. Jede Goldrhodanidlösung, die mit metallischem Golde in Gleichgewicht sein soll, muss sowohl Aurigold wie Aurogold enthalten. Zur Messung von Aurorhodanid-Gold- 1) Zeitschr. physik. Chem. 59, 118 (1907); 63, 325 (1908). Vgl. auch Zeitschr. f. Elektrochem, 17, 59 (1911). 4* 28 Potentialen ist es notwendig, solche Aurorhodanidlésungen anzuwenden, in welchen die Bildung einer kleinen Menge Aurirhodanid genügt, um die Lösung mit einer Gold- überfläche in Gleichgewicht zu bringen. Sonst zerfällt die Lösung an der Goldelektrode in Gold und Aurirhodanid in einer Ausdehnung, die störend auf die Potentialmessung wirken muss, und wohldefinierte Verhältnisse treten erst ein, wenn die ganze Lösung mit der Goldelektrode in Gleichgewicht gekommen ist, was schwierig zu erreichen sein wird. Nach der Seite 13 aufgestellten Formel für das Aurorhodanid-Aurirhodanid- Gleichgewicht an einer Goldoberfläche: [AuRh,] _ 33 [Au Rh, > ARE. OUTRE kann man berechnen, dass Gleichgewicht in den von uns angewandten Aurorhodanid- lösungen schon erreicht worden ist, wenn 0,1 bis 1°/, von dem gesamten Goldgehalt als Aurigold vorhanden ist. Es ist deshalb verständlich, dass unsere Messungen durch die Aurirhodanidbildung an der Goldelektrode nicht gestört worden sind. Die Komplexität des Dirhodanoauroatkomplexes. Mit Hilfe des jetzt bekannten Normalpotentials Gold-Aurorhodanid kann man die Beständigkeit des Aurorhodanid- komplexes mit denjenigen anderer Aurokomplexe vergleichen, fiir welche die Normal- potentiale bekannt sind. BopLANDER?*) hat durch Messung von Goldelektrodenpoten- tialen in alkalischen Aurocyanidlösungen das Normalpotential für die Reaktion: Au(CN); +6 = Au+2CN- zu — 0,611 bestimmt. Wenn man die Komplexitätskonstante des Aurorhodanids K und die des Aurocyanids K’ nennt: ER [Au Rh,] _ dnd e [Au(CN), [ [Aut] [Rh F [Aut] [CN-}?’ und wenn „E und „E’ die beiden Normalpotentiale bezeichnen, so gilt denn bei Gleichgewicht in einer Lösung müssen das Gold-Aurorhodanid- und das Gold- Aurocyanid-Potential gleich gross sein. Durch Einsetzen erhält man hieraus: ; 0,689—(—0,611) KE gg 7048 AT!) ug Die Komplexitätskonstante des Aurocyanids ist also 10% mal grösser als die des Aurorhodanids. Trotzdem die Beständigkeit des Cyanids somit weit grösser ist als die 1) Ber. deut. chem. Ges. 86, 3933 (1903). 29 des Rhodanids, kann Rhodanwasserstoff in saurer Lösung doch in nicht unbedeutendem Grade Cyanwasserstoff aus dem Cyanid austreiben; denn in saurer Lösung kommt der schwach saure Charakter des Cyanwasserstofls stark zur Geltung. Wenn man annimmt, dass Zwischenprodukte, wie Au(CN)RhT, nicht gebildet werden, kann man leicht die Gleichgewichte berechnen. Es gilt kK’ | [Au(CN),]-[Rh F K = [AuRh}-(CN p — SA In saurer Lösung, genauer bei allen Wasserstoffionenexponenten, die kleiner sind als 9,3, ist das nicht ans Gold gebundene Cyan vorzugsweise als undissoziierter Cyan- wasserstofi vorhanden; denn da die Dissoziationskonstante des Cyanwasserstoffs (K,) gleich 107% ist, hat man [HEN] [H+] 9,3 — IGN Lacets 1 FA, (2) Für saure Losungen ist es daher anschaulicher mit Cyanwasserstoffkonzentrationen statt mit Cyanionenkonzentrationen zu rechnen. Durch Elimination von [EN] zwischen (1) und (2) erhält man [Au(CN),]- [Rh]? [Au Rk, ]- [HCN — 19°81 ?PH, (3) Diese Gleichung zeigt, dass Cyanwasserstoff noch in 1 molarer Wasserstoffionenlésung (pH = 0) Rhodanionen aus ihrem Aurokomplex austreiben wird. Wenn man z. B. zu einer Lösung von Dirhodanoauroat in 1 molarer Salzsäure die äquivalente Menge Cyan- wasserstoff zufügt (2HCN pro AuRh,), werden ca. 95 % des Rhodans ausgetrieben und durch Cyan ersetzt; denn 3 aus (=) = 10° folgt tu 10:95. In Lésungen, die nicht nur stark sauer sind, sondern in welchen auch die Rhodan- ionen in grossem Uberschuss vorhanden sind, wird natiirlich der Rhodanokomplex mehr vorherrschen. Da dieses Verhalten fiir einen folgenden Abschnitt von einem gewissen Interesse sein wird, haben wir nach der Formel (3) den Zustand des Cyanwasserstoffs in verschiedenen sauren, rhodanionenhaltigen Aurorhodanidlésungen berechnet. In 1m HCl, 0,4m NaRh, 0,003 m Au! sind bei einem Gesamtinhalt von 0,0001 m Cyan 98 % des Cyans als Cyanwasserstoff und nur ca. 2 % als Aurokomplex vorhanden. Wenn der Cyaninhalt 0,001 m beträgt, sind ca. 16 % des Cyans als Aurokomplex vor- handen. In 0,1 m HCl, 0,4 m NaRh, 0,003 m Au! sind bei einem Cyaninhalt von 0,0001 m 53 % und bei einem Cyaninhalt von 0,001 m 80 % als Aurokomplex vorhanden. Die Hydrolyse des Dirhodanoauroatkomplexes. Trotz seiner im Vergleich mit dem Aurocyanidkomplex nur geringen Komplexität ist der Aurorhodanidkomplex doch sehr beständig. In Gegenwart von überschüssigen Rhodanionen ist er z. B. so wenig 30 hydrolysiert, dass man ohne Schwierigkeit die freie Säure in einer Aurorhodanidlésung mit Natriumhydroxyd titrieren kann, und zwar nicht nur mit einem Indikator wie Methyl- orange, der in saurer Lésung umschlägt, sondern auch mit Phenolphtaléin. Beim Um- schlagspunkt des Phenolphtaléins, pa = 9, ist der Aurorhodanidkomplex also in Gegen- wart von überschüssigen Rhodanionen noch nicht merkbar hydrolysiert. Dagegen wird er durch einen Uberschuss von freiem Alkali sogleich zersetzt. Der beständigere Auro- cyanidkomplex ist bekanntlich so wenig hydrolysiert, dass er nicht einmal in stark alka- lischer Lösung zersetzt wird. Ohne einen Überschuss von Rhodanionen ist der Aurorho- danidkomplex nur wenig beständig; wenn man z.B. Kaliumaurorhodanid in reinem Wasser löst, wird es schnell teilweise zersetzt. Experimentelles. 5 cm? 0,5m HAuCl, wurden mit 2g NaRh versetzt, mit kon- zentriertem Natriumsulfit entfärbt und mit 1m Natronlauge titriert; mit Methylorange wurden dabei 7,45 cm? und mit Phenolphtaléin 7,50 cm? verbraucht. Nach der folgenden Gleichung, H AuCl, + 2 NaRh+ Na,50,+3NaOH = NaAuRh, + 4 NaCl + Na,SO, + 2 H,0, berechnet man einen Verbrauch von 7,50 cm?. IV. Zusammensetzung und Komplexitat des Aurirhodanid- komplexes. Uber die Schwierigkeiten bei der direkten Messung von Aurirhodanid-Gold-Potentialen. Es liegt nahe, die Zusammensetzung und Komplexität des Aurirhodanidkomplexes durch direkte Messung von Goldelektrodepotentialen in Aurirhodanidlésungen bestimmen zu wollen. I Diese Messung lässt sich aber nur schwierig durchführen. Die meisten Aurirhodanid- lösungen wirken nämlich nach dem Reaktionschema stark lésend auf eine Goldelektrode, und erst wenn eine grosse Menge Aurorhodanid in der Lösung gebildet worden ist, tritt Gleichgewicht ein. Diese goldlösende Fähigkeit lässt sich leicht durch Wägen von dünnem Goldblech verfolgen oder durch Einhängen von schwach vergoldeten Platindrähten demonstrieren. Solange die Lösung nicht mit Gold in Gleichgewicht ist, werden die drei Potentiale, Auri-Auro, Auro-Gold, Auri-Gold, in der Lösung verschieden sein, und das mit einer Goldelektrode gemessene Potential wird ein Mittelding zwischen diesen Potentialen sein. Nur wenn die Aurirhodanidlösung eine solche Zusammensetzung besitzt, dass die Bildung einer ganz kleinen Menge von Aurorhodanid, z. B. 1 %, genügt, um die Lösung mit Gold in Gleichgewicht zu bringen, wird man durch Eintauchen einer Goldelektrode in die Lösung das Aurirhodanid-Gold-Potential direkt messen können. Die Kurve in Figur 2, Seite 14, ergibt als die hierfür notwendige Bedingung, dass die Rhodanionen- 31 konzentration nur ca. 1 % der Goldkonzentration betragen darf. Mit solchen Lösungen zu arbeiten, ist aber recht unmöglich: Erstens geht die Autoreduktion wegen der kleinen Rhodanionenkonzentration sehr schnell vonstatten, zweitens ist in einer solchen Lösung die Rhodanionenkonzentration nicht genau zu bestimmen, und drittens wird die Rho- danionenkonzentralion wegen der Bildung von Rhodanionen bei der Autoreduktion durch eine ganz geringe Autoreduktion sehr stark geändert. Die Messung des Aurirhodanid-Aurorhodanid-Potentials. Um die Zusammensetzung und Komplexität des Aurirhodanids zu bestimmen, haben wir das Auri-Auro-Potential in Mischungen von Aurirhodanid und Aurorhodanid gemessen. Ein Platinblech, das in eine gemischte Lösung von Aurorhodanid und Aurirhodanid eintaucht, besitzt ein recht wohl definiertes Potential, und eine nähere Untersuchung hat uns gezeigt, dass die potentialbestimmende Reaktion in einer solchen Lösung der reversible Übergang zwischen Aurorhodanid und Aurirhodanid ist, indem das Platin als unangreifbare Elektrode wirkt. Die Messung dieses Potentials bietet viel Interesse dar. Erstens können wir aus seiner Änderung mit der Zusammensetzung der Lösung die Formel des Rhodano- auriatkomplexes bestimmen. Zweitens können wir aus seiner absoluten Grösse, wenn das Normalpotential Aurorhodanid-Gold bekannt ist, das Gleichgewichtsverhält- nis zwischen Aurorhodanid und Aurirhodanid an einer Goldoberfläche berechnen. Drittens können wir daraus das nicht direkt messbare Potential Aurirho- danid-Gold berechnen und dadurch ein Mass für die Komplexität des Aurirho- danids erhalten. Endlich können wir, wenn wir dieses Potential kennen, aus dem Potential einer Platinelektrode in einer unbekannten Goldrhodanidlösung das Ver- hältnis zwischen Auro- und Aurirhodanid in der Lösung berechnen. Bei allen unseren Messungen wurden die Goldlösungen aus 2 molarer Salzsäure, 1 molarem Natriumrhodanid und 0,04 molarem Wasserstoflaurichlorid dargestellt. Zuerst wurde nur so viel Gold zugesetzt, als Aurorhodanid in der Lösung gewünscht wurde, dann wurde die Lösung mit Natriumsulfit genau entfärbt, darauf wurde wieder Goldlösung zugesetzt, so dass die Lösung den gewünschten Inhalt von Aurirhodanid erhielt, und endlich wurde mit Wasser auf 50 cm® verdünnt. Die Potentiale wurden gegen eine 0,1 normale Kalomelelektrode gemessen unter Einschaltung einer gesättigten Kaliumchlorid- lösung, um das Diffusionspotential zu vermindern. Die Potentialmessungen wurden mit einem Kompensationsapparat von Worrr, Berlin, einem empfindlichen Spiegelgalvano- meter von SIEMENS und HALSKE und einem Weston-Normalelement ausgeführt. Bei den ersten Messungen wurde als unangreifbare Elektrode ein in ein Glasrohr ein- geschmolzener, mit Kaliumcyanid gereinigter, blanker Platindraht benutzt. Die mit ihm gemessenen Potentialwerte zeigten sich konstant und waren sowohl unempfindlich gegen Schütteln wie wenig polarisierbar. Sie sind in Tabelle 8 zusammengestellt. Als wir später einige von den Messungen wiederholen wollten, war es unmöglich, mit blanken Platin- drähten gute Messungen zu erhalten, und wir erhielten erst gut reproduzierbare Messungen, als wir dazu übergingen, die Platindrähte zu platinieren. Aus Gründen, die wir uns nicht erklären können, erhielten wir bei allen diesen neueren Messungen Potentiale, die ungefähr 15 Millivolt höher waren als die älteren; auch Messungen mit blanken Platindrähten waren so viel höher als die früheren. In Tabelle 9 sind eine Reihe von den neueren Messungen 32 zusammengestellt. Bei den Wiederholungen sind die Messungen zu verschiedenen Zeiten und mit verschiedenen Elektroden und Goldlösungen ausgeführt. Tabelle 8. Potential von Ketten: Pt | Goldlösung | gesätt. KC1|0,1m KCl, HgCl| Hg bei ca. 22,5°. 1. Serie (ungenau). Nr. Goldlösung [Rh] E JE E #5 1. 0,002m Au!//, 0,002 m Au‘, 0,1m NaRh, 1m HCl 0,088 0,3410 —-0,0005 0,3570 0,2951 220,002 ..— NON r= 0000 fo eed) eee = 0,3375 0,3535 0,2916 3. 0027 SOUR ees a Sneed ERR 0,188 0,3217 —0,0003 0,3377 0,2946 BE IE UEES Gage ta IE ee 0,388 0,2971 —0,0001 0,3131 0,2890 Sunrise 2007 igi Eee 0,076 0,3421 —0,0004 0,3581 0,2926 6°) 0,002) E-54.[0006 ‘SA rede 0,180 0,3094 —0,0002 0,3254 0,2958 A.) BON te AO ODS 4 DO at 0,176 0,3200 —0,0000 0,3360 0,2918 8: :0.006, a=. 50-0026 100 ee seer 0,172 0,3365 —0,0012 0,3525 0,2937 Mittel: 0,2930 Tabelle 9. Potential von Ketten: Pt|Goldlôsung | gesätt. KC1| 0,1 m KCl, HgCl| Hg 1) Bei 18°: Goldlösung [Rh] E JE E E, 0,002 m Au!” 0,002 m Alf, 0,1 mNaRh, 1m HCl 0,088 0,3557 —0,0004 0,3717 0,3108 0,3343 —0,0002 0,3503 0,3085 | 0,3346 —0,0001 0,3506 0,3088 0.002 == DO 6 9.102 eae 0,188 0,3347 —0,0008 0,3507 0,3089 | 0,3344 —0,0010 0,3504 . 0,3086 0,3348 —0,0010 0,3508 0,3090 0,00. EN oor MO UE tee es 0,176 0,3338 —0,0004 0,3498 0,3063 Mittel: 0,3086 2) Bei 8,5°: 0,002 m Au”, 0,002 m Au!, 02m NaRh, 1m HCL 0,188 { oe a Es Mittel: 0,3064 Bei der Angabe der Zusammensetzung der Lösungen bedeutet Au’ Aurigold und Au’ Aurogold. [Rh”] bezeichnet die Rhodanionenkonzentration der Lösungen; sie ist berechnet aus der angewandten Menge Natriumrhodanid unter der Annahme, dass das Aurigold als Tetrarhodanokomplex und das Aurogold als Dirhodanokomplex vorhanden ist. Eist das einige Minuten nach der Darstellung der Lösung gemessene Potential, und JE ist die Änderung des Potentials in den nächsten 10 Minuten. E’ ist das für das Dif- fusionspotential korrigierte Potential und ist berechnet aus E durch Addition von 0,0160%). o£ ist endlich das nach der folgenden Formel berechnete Normalpotential: ’ BR EN [Auf] nee A F Fe [Au]. [Rh ]* Die Zusammensetzung des Aurirhodanidkomplexes. Die Werte, die in den Tabel- len 8 und 9 für das Normalpotential berechnet sind, weisen innerhalb jeder Messungs- 1) Vgl. Seite 27. 33 reihe eine gute Übereinstimmung auf. Wir können daraus schliessen, dass die potential- bestimmende Reaktion die folgende sein muss: Aurokomplex + 2Rh~ = Aurikomplex + 26; denn nur diese Reaktion führt zu der bei der Berechnung des Normalpotentials benutzten Formel. Da der Aurokomplex AuRh, ist, muss folglich der Aurikomplex Au Rh7 sein. Um einen Begriff von der Genauigkeit zu geben, mit welcher man auf die Formel AuRhz schliessen darf, kann Folgendes angeführt werden. Wenn E, und E, die bei einer bestimmten Goldkonzentration bei zwei verschiedenen Rhodanionenkonzentrationen ge- fundenen Potentiale sind, kann man nach einer Formel, die den früher benutzten!) Bop- LANDER’schen analog ist, berechnen, wie viel Rhodanradikale der Aurikomplex mehr enthält als der Aurokomplex. Wenn diese Zahl n genannt wird, gilt nämlich: ae Le (RT:2F)(In[Rh |], —In{Rh],) Wenn man aus Tabelle 8 das Mittel von Versuch 1 und 2 nimmt und mit Versuch 3 kombiniert, findet man n — 1,82; wenn man mit Versuch 4 kombiniert, findet man n = 2,23; und wenn endlich Versuch 5 mit 7 kombiniert wird, erhält man 2,07. Wenn man eine saure, aurirhodanidhaltige Lésung mit Ather ausschiittelt, nimmt die Atherschicht eine stark rotbraune Farbe an, und die wässrige Schicht wird vollstandig entfarbt. Die roten Rhodanoaurikomplexe lassen sich also mit Ather vollstandig extra- hieren. Es liegt nahe, zu versuchen, die Zusammensetzung der roten Komplexe durch Atherextraktion zu bestimmen; eine solche Bestimmung lässt sich indessen nur schwierig durchführen, weil Rhodanwasserstoff auch durch Äther ausgeschüttelt wird. Wenn man die Ätherschicht mit säurehaltigem Wasser wäscht, kann man zwar den Rhodanwasserstoff entfernen; gleichzeitig wird aber durch Dissoziation oder Autoreduk- tion neuer Rhodanwasserstoff gebildet; wenn man das Waschen fortsetzt, fangt zuletzt eine tiefergreifende Zersetzung an, und die Atherschicht wird getriibt. In der Atherschicht - ist die Zahl der Rhodangruppen pro Goldatom dann gewohnlich bis unter drei gesunken, aber trotzdem ergibt das Waschwasser immer Rhodanreaktion mit Ferrisalz. Das Normalpotential Aurorhodanid-Aurirhodanid. Für das Normalpotential findet man als Mittel aus unseren ersten Versuchen 0,2930 Volt, während das Mittel aus allen unseren späteren Versuchen 0,3086 ist. Da letzterer Wert zu verschiedenen Zeiten und mit verschiedenen Elektroden immer wieder gefunden worden ist, muss er als der richtigere betrachtet werden. Wir werden im Folgenden mit dem Werte 0,308 rechnen. Wenn dieser Wert durch Addition von 0,337 von der Zehntelnormal-Kalomelelek- trode auf die Normal-Wasserstoffelektrode umgerechnet wird, erhält man „Er = 0,645. Für die Reaktion J, (fest) + 20 = 2 J” ist das Normalpotential „Er = 0,54, und für die Reaktion Br, (flüssig) + 20 = 2 Br” ist das Normalpotential „Er = 1,09. Das Auri- rhodanid ist hiernach als Oxydationsmittel etwas stärker als Jod, aber viel schwächer als Brom. 1) Seite 24. D, K. D. Vidensk. Selsk. Skr., naturvidensk. og mathem. Afd., 8. Række, V. 1. 5 34 Die Berechnung des Aurirhodanid-Gold-Potentials. Aus den zwei Normalpotentialen Aurirhodanid-Aurorhodanid und Aurorhodanid-Ggld kann man das Normalpotential Aurirhodanid-Gold berechnen; denn die elektrische Energie, die man reversibel durch Abscheidung des Goldes aus Aurirhodanid in einem galvanischen Element erhalt, muss der Summe der elektrischen Energien gleich sein, die man erhält, wenn man reversibel in galvanischen Elementen erst das Aurirhodanid in Aurorhodanid und danach das Auro- rhodanid in Gold umwandelt. Wenn man die Potentiale der drei Prozesse in derselben Lösung mit E(AuRh,, Au), E(AuRhz, AuRh,), E (AuRh;, Au) bezeichnet, muss daher gelten: 3E (AuRkh,, Au) = 2E(AuRh,, AuRh,) + E(AuRh;,, Au)!). Wenn man diese Gleichung auf eine Lösung anwendet, die in betreff der verschie- denen Ionenarten 1 molar ist, werden die E-Werte zu den Normalpotentialen, und man erhält deshalb für das Normalpotential Aurirhodanid-Gold: 3 fk = 2- 0,645 + 0,689; „En = 0,660. Die Komplexität des Aurirhodanids. Mit Hilfe des jetzt bekannten Normalpoten- tials Gold-Aurirhodanid können wir die Komplexität des Tetrarhodanoauriatkom- plexes mit denjenigen anderer Aurikomplexe vergleichen, für welche die Normalpoten- tiale bekannt sind. Da bisher keine solche Normalpotentiale bekannt waren, haben wir das Normalpotential Gold-Tetrachloroauriatkomplex bestimmt; es ist gleich 1,0012). Wenn man die beiden Komplexitätskonstanten X und K’ nennt, [Au Rh, ] K' — [Au CI, ] [Aut**]. [Rh]? [Aut**). [CT ’ und wenn „E und ,E' die beiden Normalpotentiale bezeichnen, so gilt: É 3 a E-ıE = 3F In K° da bei Gleichgewicht in einer chloridhaltigen und rhodanidhaltigen Aurilésung das Gold-Aurichlorid-Potential und das Gold-Aurirhodanid-Potential gleich gross sein müssen. Durch Einsetzen erhält man hieraus: ; 1,001 — 0,660 K “tan MODE, oe 102707 Die Komplexitätskonstante des Tetrarhodanoauriatkomplexes ist also ca. 1017” mal grösser als die des Tetrachloroauriatkomplexes, 1) Die Lurner’sche Regel. Vgl. Luruer, Zeitschr. physik. Chem. 34, 488 (1900); 86, 385 (1901). *) Vgl. den Anhang dieser Abhandlung. 35 V. Die Dissoziation des Aurirhodanids in Aurorhodanid und freies Rhodan. Nachdem wir gefunden hatten, dass das Auri-Auro-Potential in einer Goldrhodanid- lösung an einer Platinelektrode sich schnell einstellt, lag es nahe, die Autoreduktion in einer Aurirhodanidlösung durch Potentialmessungen mit Platinelektroden zu verfolgen. Die Ausführung von solchen Messungen führte uns zur Entdeckung einer unerwarteten Dissoziation von Aurirhodanid in Aurorhodanid und Rhodan. Messungen von Platinelektrodenpotentialen in Aurirhodanidlösungen. Die Messungen wurden in ähnlicher Weise wie die im vorigen Abschnitt besprochenen ausgeführt. Die gewonnenen Resultate sind in den Tabellen 10 und 11 enthalten. In den Tabellen bedeutet E das Potential, das schnell (1 bis 3 Minuten) nach der Darstellung der Gold- lösung gefunden wurde. Ausserdem sind einige Angaben über die zeitliche Änderung des Potentials mitgeteilt. Tabelle 10. Potential von Ketten: (Pt blank) | Aurirhodanid | gesätt. KC1|0,1 m KCl, HgCl| Hg bei 21,5°. Aurirhodanidlösung E in aus 35 = Min. 0,004 m HAuCl,, 0,4 m NaRh, 1 m HCl 0,3260 0,0006 0,0013 0,002 _- 0,4 — 1 — 0,3230 0,0052 0,004 = 0,2 — 1 — 0,3450 0,0007 0,002 — 0.2 — 1 — 0,3362 0,0014 Anm. Gleichzeitig wurde mit derselben Apparatur das Normalpotential Aurirhodanid-Aurorho- danid zu 0,2930 gegen die Dezinormal-Kalomelelektrode bestimmt. Tabelve. ti Potential von Ketten: (Pt platiniert) | Aurirhodanid | gesätt. KCl|0,1 m KCl, HgCl\ Hg bei 18°. Abnahme von E Aurirhodanidlösung E in 10Min. in 75Min. in 150 Min. 0,004 m HAuCl,, 0,2 m NaRh, 1 m HCl 0,3625 0,0012 0,0090 0,002 — ee EEE 0,3534 0,0010 0,0020 0,001 — aber 0,3542 0,0083 0,002 — REISEN = 0,368!) 0,015 0,002 — dt at 0,3771 0,0041 Anm. Gleichzeitig wurde mit derselben Apparatur das Normalpotential Aurirhodanid-Aurorho- danid zu 0,3086 gegen die Dezinormal-Kalomelelektrode bestimmt. Aus den Potentialmessungen in den Tabellen 10 und 11 haben wir die Aurorhodanid- konzentrationen in den verschiedenen Aurirhodanidlésungen berechnet. Die Resultate dieser Berechnungen sind in den Tabellen 12 und 13 mitgeteilt. Zuerst sind die gemessenen Potentiale für das Diffusionspotential korrigiert. Die nach der HENDERSON’schen Formel?) berechneten Werte des Diffusionspotentials sind in den Tabellen angegeben. Aus E’, dem korrigierten Potential, ist dann das Verhältnis zwischen Aurirhodanid und Aurorhodanid nach der folgenden Gleichung berechnet: 1) Dieser Wert ist wegen der schnellen Abnahme des Potentials ein auf die Zeit 0 extrapolierter Wert. ?) Vgl. Seite 27 dieser Abhandlung. 5* 36 ‚ch RT [Au Rk] 2 0 ar" OE ss [BETTS Für ,E, das Normalpotential Aurorhodanid-Aurirhodanid, sind die in den Anmer kungen zu den Tabellen 10 und 11 mitgeteilten Werte angewandt. Zuletzt ist der Disso- ziationsgrad des Aurirhodanids, a, nach der folgenden Gleichung berechnet: 1—a _ [AuRh,] a a [Au Rhe |" Tabelle 12. Der Dissoziationsgrad des Aurirhodanids, elektrometrisch bestimmt. Berechnet aus der 1. Messungsreihe (mit blanker Platinelektrode und bei 21,5°). ; u 3 4 Aurorhodanidgehalt (100 a) Zusammensetzung der Lösung Dif. pot. [Rh] sogleich nach 10 Min. nach 60 Min. 0,004 m HAuCl, 0,4 m NaRh, 1m HCl — 0,016 0,384 12,4 lo 12,9 lo 13,6 lo 0,002 — 0,4 == 1 — — 0,016 0,392 14,7 — nn. = 20,6 — 0,004 — 0,2 — 1 — — 0,016 0,184 11,9 — 12,5 — ae 0,002 — 0,2 — 1 — — 0,016 0,192 20,5 — 23,9 — a Tabelle 13. Der Dissoziationsgrad des Aurirhodanids, elektrometrisch bestimmt. Berechnet aus der 2. Messungsreihe (mit platinierter Platinelektrode und bei 18°). Zusammensetzung der Lösung Dif. pot. [Rh |] sogl Aurora Fa SE Min 0,004 m HAuCl,, 0,2 m NaRh, 1 m HCI — 0,016 0,186 9,9 lo 10,7 °o 18,3 "lo 0,002 — 0,2 — 1 — — 0,016 0,192 12,6 — 13,3 — 26,3 — 0,001 — C2. 1 = — 0,016 0,196 16,1 — se See = 0,002 — 0,2 — 0,2 — — 0,007 0,192 12,4 — 31,0 — te 0,002 — 0,1 — 1 — — 0,016 0,092 12,4 — 16,4 — su — Wir sehen aus den Tabellen 12 und 13, dass Aurirhodanidlésungen gleich nach ihrer Darstellung eine recht bedeutende Menge Aurogold enthalten, und dass die Menge von Aurorhodanid nur langsam grösser wird. Diese Resultate stimmen anscheinend gar nicht mit den Resultaten überein, die man durch Sulfittitrierungen erhält; denn danach scheint in den drei Minuten alten Lösungen noch alles Gold als Aurirhodanid vorhanden zu sein (vgl. die später beschriebenen, zur Kontrolle der Sulfittitrierung ausgeführten Titrie- rungen). Wir können daraus schliessen, dass das in den frischen Aurirhodanidlösungen elektrometrisch nachgewiesene Aurorhodanid wahrscheinlicherweise durch einen um- : kehrbaren Dissoziationsprozess entstanden sein muss; in diesem Falle wird im Laufe der Sulfittitrierung das durch die Dissoziation gebildete Aurorhodanid in Aurirhodanid zurückgewandelt werden, und es ist deshalb verständlich, dass alles Gold als Aurirhoda- nid titriert wird. Das sogleich gebildete Aurorhodanid kann nicht durch Autoreduktion gebildet sein, denn dieser Vorgang ist bekanntlich nicht reversibel. Schon die Beobach- tung, dass 10 bis 20 % Aurorhodanid momentan gebildet werden, und dass eine weitere Menge Aurorhodanid nur langsam gebildet wird, deutet übrigens darauf, dass ein sich schnell einstellendes Dissoziationsgleichgewicht über die langsame, irreversible Auto- reduktion gelagert hat. Um die Richtigkeit dieser Dissoziationshypothese zu prüfen, haben wir eine Reihe von Messungen über die Farbenintensität von Aurirhodanidlösungen ohne und mit Aurorhodanidzusatz ausgeführt. we Er 37 Spektrophotometrische Untersuchungen über die Farbe von Aurirhodanidlösungen. Um für die Farbe des Aurirhodanids ein Mass zu erhalten, haben wir den molaren Extink- tionskoeffizienten des Goldes in Aurirhodanidlösungen für die Wellenlänge 578 py bestimmt. Der molare Extinktionskoeffizient, mEk, ist in folgender Weise definiert: meth 77) mou; mEk = wo I die Schichtdicke in Zentimetern, C die molare Konzentration des Goldes, J, die Intensität des einfallenden Lichtes und / die Intensität des austretenden Lichtes ist. Die Bestimmung des molaren Extinktionskoeffizienten wurde mit einem Könıc- MARTENnS’schen Spektrophotometer von SCHMIDT und Haenscu ausgeführt. Der Apparat war mit RUTHERFORDprisme versehen, und als Lichtquelle wurde eine Quarz-Quecksilber- lampe benutzt. Die Absorption des Aurirhodanids wächst stark gegen das violette Ende des Spektrums. Bei 546 zu» (der grünen Quecksilberlinie) ist der molare Extinktionskoef- fizient ca. 2,6mal grösser als bei 578 zu (der gelben Linie), und bei 436 zu (der blauen Linie) ist er fast 10mal grösser. Wir benutzten die Wellenlänge 578 yu, weil die Lösungen, die wir zu messen wünschten, hier bei bequemen Schichtdicken eine Extinktion von pas- sender Grösse besassen. Die untersuchten Lösungen wurden aus 5m Salzsäure, 1,4 m Natriumrhodanid, Wasser und 0,05 m Wasserstoflaurichlorid zusammengemischt. Ein- zelne Lösungen wurden doch aus 2 m Natriumrhodanid dargestellt und die konzentrierte- sten durch Abwägen von festem Natriumrhodanid. Wenn die Lösung ausser Aurirhodanid auch Aurorhodanid enthalten sollte, wurde das Wasserstoflaurichlorid in zwei Portionen zugesetzt, und die erste Portion wurde sorgfältig mit der eben notwendigen Menge von Natriumsulfit reduziert, ehe die zweite zugesetzt wurde. Das Alter der Lösungen wurde vom letzten Zusatz von Wasserstoflaurichlorid an gerechnet. Die Konzentration des mit Sulfit reduzierten Goldchlorids wird als Au? und die Konzentration des nicht reduzierten als Au?!’ angegeben. In den folgenden Tabellen ist das experimentelle Material zusammen- gestellt. Tabelle 14. 1 m HCl, 0,28 m NaRh. 0,002 m AulII 0,002 m AulII 0,005 m Aufl! 0,005 m Aufl! 0,005 m AulII 0,005 m Au! 0,005 m Auf 0,010 m Auf 16° 16° 23° 22,5° 23° Alter m Ek Alter mEk Alter mEk Alter mEk Alter m Ek O Min. (126) OMin. (145) OMin. (131) OMin. (142) OMin. (143) 3 — 126,1 3 — 144,3 3) — 130,0 4,5— 141,6 4 — 142,2 8,5— 125,8 36 — 142,5 15 — 127,2 82 — 133,3 16 — 140,8 18 — 125,5 87 — 140,2 35 — 124,8 268 — 123,4 60 — 138,9 40 — 124,6 1230 — 127,6 154 — 115,2 480 — 117,5 129 — 135,9 75 — 122,9 — — — 266 — 110,0 1310 — 102,7 447 — 131,3 126 — 121,1 — — — 1370 — 83,4 — — — 1310 — (114,5)! 2007 — 99,1 — — _ 5760 — 18,8 — — — — — — Sulfittitrierung: Sulfittitrierung: 1210 Min. 73°! AulII 1240 Min. 90°/o Aul1II 1) Die Lösung hatte Nadeln von NaAuRh, ausgeschieden. 38 Tabelle 15. 1m HCl, 0,14 m NaRh. 0,002 m Au!!! 0,002 m AuZZZ 0,005 m AuZZZ 0,005 m Au!!! 0,005 m Auf! 0,005 m Au? 20°—18° 18° 21°—18° 18° Alter mEk Alter mEk Alter mEk Alter mEk 0 Min. (111) 0 Min. (126) 0 Min. (115) 0 Min, (124) 3 — 107 9 3 — 126,1 3 — 111,7 3 — 123,4 8 — 104,9 17 — 125,5 12,5— 107,5 13 — 122,9 32 — 98,0 32 — 125,6 37,5 — 102,0 129 — 121,2 70 — 92,9 153 — 124,7 60 — 98,4 234 — 118,9 105 — 89,7 — — — 84 — 96,6 — — — 1240 — 59,9 — — — 121 — 92,4 — — — — — — — — — 159 — 90,6 — — _ = = — — — — 1220 — 68,8 — = — Sulfittitrierung : Sulfittitrierung : Sulfittitrierung: Sulfittitrierung: i 1250 Min. 51% Au? 160Min. 100% Au! 1230 Min. 55° Au!!! 240Min. 98,2% Au!" | Tabelle 16: 1 m HCl, 0,056 m NaRh. É 0,002 m Auf’! 0,002 m Auf! 0,005 m Au!" 0,005 m Auf!!! 1 0,005 m Au? 0,005 m Au! i 189 18? 18° 18° Alter mEk Alter mEk Alter mEk Alter mEk 0 Min. (99) 0 Min. (112) 0 Min. (100) 0 Min. (109) 3 — 88,4 3 — 111,8 3 — 86,6 3 — 107,6 8 — 80,1 9 — 1143 T — 79,4 16 — 102,6 16 — 71,9 82 — 107,4 15 — 71,2 38 — 96,8 25 — 66,7 — — — 54 — 56,4 1325 — 52,1 36 — 62,6 — — — 82 — 52,4 — — = 57 — 55,7 — — — 1180 — 26,1 — — — 70 — 53,0 = — = ua 2 EN WET AS - 108 — 47,5 — — — — — — — — — 143 — 43,7 — — — — — — — = — 1196 — 22,9 — — — — = i — — — — Sulfittitrierung: Sulfittitrierung: Sulfittitrierung: Sulfittitrierung: 1200 Min. 24° Au? 90Min. 97% Auf! 1200 Min. 26/0 Aull! 1330 Min. 48/9 Auf’! Tabelle 17. Tabelle 18. 1 m HCl, 0,023 m NaRh 0,002 m Aul!1 1m HCl, 0,8 m NaRh 0,002 m Aul11 0,002 m Auf!! 0,005 m Auf? 0,005 m Au! 15° 16,5° 16,5° Alter mEk Alter m Ek Alter mEk 0 Min (102) 0 Min. (167) 0 Min. (182) 3 — 100,8 3 — 167,2 3, — 181,3 13 — 96,8 9 — 167,0 19 — 179,6!) 30 — 91,8 16 — 166,6 — — — 69 — 82,4 66 — 165,67) — — — 99 — 78,0 20 — 63,2 350 — 52,5 2895 — 16,0 ') Die Lösungen wurden schnell unklar und hatten am nächsten Tage einen gelben Niederschlag ausgeschieden. 39 Tabelle 19. Tabelle 20. 0,5 m HCl, 0,28 m NaRh 0,2 m HCI, 0,28 m Na Rh 0,005 m Au!!! 0,005 m Au!!! 0,005 m Auf?! 0,005 m Auf?! 0,005 m Au? 0,005 m Auf 22,5°-—-20° 22,5°—20° 16° 16° Alter mEk Alter mEk Alter mEk Alter mEk 0 Min. " (130) 0 Min. (137) 0 Min. (124) 0 Min. (135) 3,5 — 127,1 9,5 — 135,2 2,5 — 119,3 3 — 134,3 21,5 — 120,0 12,5 — 134,8 6 — 114,4 7 — 133,6 46 — 114,7 37 — (131,8) 12 — 107,8 16 — 132,8 111 — 105,8 60 — (125,8) 33 — 96,5 33 — (129,9) — — _ 128 — (113,3) ') 58 — 89,8 49 — (125,3) 106 — 80,6?) 74 — (119,5) — — — 126 — (107,4)°) Tabelle 21. Tabelle 22. 0,2 m HCl, 2m NaRh 0,2 m HCl, 4m NaRh 0,001 m Au!/? 0,001 m Auf! 16° 16° Alter m Ek Alter m Ek 0 Min. (219) 0 Min. (282) 3,5 — 218 4 — 282 4) 10 — 217 33 — 214 68 — 214 145 — 209 8) Die Extinktionswerte fiir die Zeit Null wurden durch Extrapolation bestimmt und sind deshalb in den Tabellen in Parenthese gesetzt. Die Extrapolation ist doch recht sicher; nur in den rhodanarmen Lösungen ohne Aurorhodanidzusatz schätzen wir, dass die Extrapolation nicht auf 1 % sicher ist. Unter den übrigen Extinktionswerten sind weiter einige in Parenthese gesetzt, für welche wir wissen, bzw. schätzen, dass sie durch Aus- scheidung von Natriumaurirhodanid aus der Lösung entstellt sind. Von den untersuchten Lösungen wurden viele nach Abschluss der Extinktionsmessungen sulfittitriert, und die dadurch gewonnenen Resultate sind unten in den Tabellen aufgeführt. Angenäherte Berechnung des Dissoziationsgrades aus den Farbenmes- sungen. In Tabelle 23 sind die Anfangswerte des molaren Extinktionskoeffizienten für die wichtigsten untersuchten Lösungen zusammengestellt. !) Die Lösung hatte Nadeln von NaAuRh, ausgeschieden. 2) Die Lösung hatte am nächsten Tage Nadeln von NaAuRh, ausgeschieden. 3) Die Lösung hatte am nächsten Tage einen gelben Niederschlag ausgeschieden und besass noch eine rotbraune Farbe. 4) Die Lösung wurde schnell sehr unklar und hatte am nächsten Tage einen gelben Niederschlag ausgeschieden und war ganz entfärbt. 40 Tabelle 23. Anfangswerte fiir den molaren Extinktionskoeffizienten. 0,002 m Au!!! 0,002m Au! 0,005m Au!!! 0,005m Au!!! 0,005 m Au!!! 0,005 m Auf 0,005m Au! 0,010m Aut 0,056 m NaRh, 1 m HCl 99 112 100 109 AN Da a 111 126 115 124 ae 0,28 ie Gl 126 145 131 142 143 0,8 + iy 167 182 1 er er 0280 Den pen re 130 137 DSL 7 ee #3 ls 124 135 Die Tabelle zeigt, dass die Farbe des Aurirhodanids in einer Lösung ohne Aurorho- danid immer kleiner ist als in derselben Lösung mit Aurorhodanid. Da das Aurorhodanid selbst ohne Farbe ist, dürfen wir hierin eine wertvolle Stütze für die Hypothese sehen, nach welcher Aurirhodanid teilweise zu Aurorhodanid dissoziiert ist. Der Zusatz von Aurorhodanid verstärkt die Farbe, indem er die Dissoziation des Aurirhodanids zurück- drangt. Ein Zusatz von 0,005 molarem Aurorhodanid geniigt augenscheinlich zur fast vollständigen Zurückdrängung der Dissoziation; denn ein weiterer Zusatz von 0,005 molarem Aurorhodanid hat nur einen geringen Einfluss. Wenn man annimmt, dass 0,005 molares Aurorhodanid die Dissoziation vollständig zurückdrängt, kann man aus den Extinktionsmessungen die Dissoziation des Aurirhodanids nach der folgenden Formel berechnen: mk, — mEk, .100. mEk, Hier bedeutet 4 Dissoziationsgrad in Prozent, und mEk, und mEk, sind die Extink- tionskoeffizienten der Lösung bzw. mit und ohne Aurorhodanid. In Tabelle 24 sind die nach dieser Formel berechneten Dissoziationsgrade angegeben. Tabelle 24. Die Dissoziation des Aurirhodanids, spektrophotometrisch bestimmt. 1. Annäherung. 0,002 m Au!" 0,005 m Au! 0,056 m NaRh, 1m HCl 11,6 °lo 8,2 “lo 0,14 — 1 — il 7,3 - 0,28 — 1 — 13,1 - 7,7 - 0,28 — 0,5 — DRS 5,1 - 0,28 — 0,2 — IPE 8,1 - 0,8 — 1 — (8,2 - ) As Mittelwerte: 12,2 lo 7,2 lo Die Tabelle zeigt, dass die Dissoziation des Aurirhodanids von dem Rhodanionen- gehalt und von dem Wasserstoffionengehalt recht unabhängig ist. Nur wenn der Rho- danionengehalt bis 0,8 steigt, macht sich eine abnehmende Dissöziation bemerkbar; sie steht damit in Verbindung, dass in dieser Lösung das Aurirhodanid ausser als Tetrarho- danokomplex auch als Penta- und Hexarhodanokomplex vorhanden ist (vgl. später). Dagegen ändert sich der Dissoziationsgrad stark mit der Aurirhodanidkonzentration, und zwar ungefähr nach dem Osrtwap’schen Verdünnungsgesetz: 41 In 0,002 molaren Aurirhodanidlésungen war die Dissoziation durchschnittlich 12,2 % (den Versuch mit 0,8m NaRh nicht mitgerechnet). Daraus berechnet sich die Dissoziation einer 0,005 molaren Lösung nach dem Verdünnungsgesetz zu 7,9 %, während im Mittel 7,2 % gefunden wurden. Bestimmung der Dissoziationsprodukte. Aus den für den Dissoziationsgrad gefundenen Gesetzmässigkeiten können wir den wichtigen Schluss ziehen, dass das zweite Dissoziationsprodukt des Aurirhodanids das freie Dirhodan sein muss (oder eventuell ein Hydrat davon), und dass die Dissoziation also nach der folgenden Gleichung vor sich geht: AuRh, = AuRh, + Rh,. Da die Dissoziation von der Rhodanionen- und der Wasserstoffionenkonzentration näm- lich nicht beeinflusst wird, können Rhodanionen und Wasserstoffionen bei dem Dis- soziationsprozess weder gebildet noch verbraucht werden, und da die Dissoziation dem Ostwap’schen Verdünnungsgesetz folgt, werden bei ihr für jedes dissoziierte Molekül nur zwei neue Moleküle gebildet. Da das zu dissoziierende Molekül Au Rh; ist, und da das eine Dissoziationsprodukt Au Rh, ist, muss das zweite Dissoziationsprodukt die Formel Rh, besitzen. Nur die Möglichkeit liegt noch vor, dass etwas Wasser bei der Dissoziation aufgenommen wird, und dass also ein Hydrat von Dirhodan gebildet wird. Genaue Berechnung des Dissoziationsgrades. Bei der für Tabelle 24 ange- wandten Berechnung des Dissoziationsgrades haben wir angenommen, dass 0,005 molares Aurorhodanid die Dissoziation vollständig zurückdrängt, was nicht ganz exakt ist. Wenn man mittels des Verdünnungsgesetzes für die unvollständige Zurückdrängung der Disso- ziation durch 0,005 molares Aurorhodanid korrigiert, und wenn man berücksichtigt, dass die aurorhodanidhaltigen Lösungen einen etwas kleineren Gehalt an Rhodanionen be- sitzen als die entsprechenden aurorhodanidfreien, und dass ihr Extinktionskoeffizient dadurch ein wenig verkleinert wird (siehe Seite 43), erhält man die folgenden exakteren Werte für den Dissoziationsgrad. Tabelle 25. Die Dissoziation des Aurirhodanids, spektrophotometrisch bestimmt. 2. Annäherung. . 0,002 m Auf!!! 0,005 m Au!!! 0,056 m NaRh, 1m HCl 14,2 lo (18°) 11,7°%o (18°) 0,14 — 1 — 14,4 - (209) 97 = (219) 0,28 — 1 — 14,7 - (169) 9,4 - (23°) 0,28 — 0,5 — Lu ME 6,8 - (22,59) 0,28 — 0,2 — This 9,8 - (16°) Mittelwerte: 14,4 Jo 9,5 lo Aus den in der Tabelle angegebenen Mittelwerten des Dissoziationsgrades berechnet man für die Dissoziationskonstante, K,, die Werte 0,485 - 10°* und 0,498 : 10°*. Man kann also setzen: D. K. D. Vidensk. Selsk. Skr., naturvidensk. og mathem, Afd., 8. Række, V, 1. 6 = 42 a 1—a Vergleich der elektrometrisch und der spektrophotometrisch gefun- denen Dissoziationsgrade. In Tabelle 26 sind die elektrometrisch bestimmten Dissoziationsgrade (Tabelle 12 und 13, Seite 36) mit denjenigen Dissoziationsgraden zu- sammengestellt, die man aus dem spektrophotometrisch bestimmten Wert der Disso- ziationskonstante 0,49 - 10” berechnet. Ka — C: — 0,49 : 10 * bei ca. 18°. Tabelle 26. Vergleich der elektrometrisch und spektrophotometrisch gefundenen Dissoziationsgrade. elektrometrisch spektrophotometrisch I II 0,004 m Au!!!, 0,4 m NaRh, 1m HCl 12,4 Io Sr} 10.5 °| 0.004 NE mon ae 11,9 - 9,9 lo J Te 0,002 — 0,4 — 1 — 14,7 - UNE 0,002 — 0,2 — 1 — 20,5 - 12,6 - 145 - DOS NOT, EU ee 12,4 - 4 0,002 — 0,2 - 0,2 — rn 12,4 - 0,001 — 0,2 — 1 — ne 16,1 - 19,8 - Die Übereinstimmung ist so gut, wie man in Anbetracht der Genauigkeit der zwei Methoden nur erwarten darf. VI. Über die Existenz von anderen Aurirhodanid- komplexen neben dem Tetrarhodanoauriatkomplexe. . Die Bildung von Pentarhodanoauriatkomplexen und Hexarhodanoauriatkomplexen bei hohen Rhodanionenkonzentrationen. Aus unseren vielen spektrophotometrischen Mes- sungen, die hauptsächlich zur Bestimmung der Dissoziation des Aurirhodanids in Auro- rhodanid und freiem Rhodan ausgeführt wurden, kann man über den Zustand des nicht dissoziierten Aurirhodanids in den Lösungen Schlüsse ziehen. Es wurde früher durch Potentialmessungen in Mischungen von Auri- und Aurorhodanid gezeigt, dass das Auri- rhodanid in salzsauren Lösungen mit einem Gehalt von 0,1 bis 0,4m Rhodanionen vor- zugsweise als Tetrarhodanoauriatkomplex vorhanden ist. Die Potentialmessungen schlies- sen aber nicht aus, dass kleinere Mengen, z. B. 10 %, in Form von rhodanärmeren oder rhodanreicheren Komplexen vorhanden sind. Wenn solche Komplexe aber in nicht ganz verschwindender Menge vorhanden wären, so müsste die Farbe, d.h. der Extinktions- koeffizient, des Aurirhodanids sich mit der Rhodanionenkonzentration ändern; denn mit der Rhodanionenkonzentration muss sich das Mengenverhältnis zwischen den verschie- denen Komplexen ändern, und es ist ganz unwahrscheinlich, dass ihre Farben nicht ver- schieden sein sollten. In der Tabelle 27 ist der molare Extinktionskoeffizient des Aurirho- danids für A = 578 up für eine Reihe von Lösungen mit wachsender Rhodanionenkon- zentration angegeben. 43 Tabelle 27. ; Die Anderung des molaren Extinktionskoeffizienten des Aurirhodanids mit der Rhodanionenkonzentration. Temp. [Rh | mEk 0,002 m Au!/4, 0,005 m Auf, 0,028 m NaRh, 1 m HCl 15° 0,010 102 ae 0006 — 100667 = A QE 18° 0,026 109 BE 70,005 7 200667 N 18° 0,038 112 00: — 000 — 0,14 AT it. 18° 0,110 124 DO — 000 —. (0,14 EN me 18° 0,122 126 0005 — 000 — 0,28 CRE ae 22,5° 0,250 142 Ko — 0,005 — 0,8 =" Bie 16° 0,262 145 tiene — 000° — 08 ON ee 16,5° 0,782 182 Co — 0000 — 20 = We 0e 16° 2,0 219 RUN 110,000. — 40 02, #16? 4,0 282 Alle die angewandten Lösungen mit Ausnahme der zwei letzten enthalten 0,005 m Aurorhodanid, und die Dissoziation des Aurirhodanids zu Aurorhodanid und Rhodan ist daher nur in den zwei letzten von Bedeutung. Die angegebenen Werte des molaren Extinktionskoeffizienten gelten für den Zeitpunkt der Darstellung der Lösungen und sind den Tabellen 14—23 entnommen. Die Rhodanionenkonzentrationen, [RAT], sind durch Subtraktion des zur Bildung von Tetrarhodanoauriat und Dirhodanoauroat not- wendigen Rhodans von den Konzentrationen des Natriumrhodanids berechnet. Die Tabelle zeigt, dass der molare Extinktionskoeffizient des Aurirhodanids sich mit der Rhodanionenkonzentration sehr bedeutend ändert, nämlich von 102 bei 0 01 m Rh~ bis 282 bei 4,0 m Rh”. Das Aurirhodanid ist also in diesen Lösungen nicht in Form eines einzelnen, unveränderlichen Komplexes vorhanden. Wenn man die Änderung des Extinktionskoeffizienten näher studiert, wird man zu dem Resultate geführt, dass der Tetrarhodanokomplex bei steigender Rhodanionenkonzentration immer mehr und mehr in intensiver gefärbte Penta- und Hexarhodanokomplexe umgewandelt wird. In 0,14 molarem Natriumrhodanid ist der Extinktionskoeffizient des Aurirhodanids 125, und durch Steigerung der Rhodanmenge kann man ihn bis auf 282 vergrössern. Diese grosse Steigerung kann nicht durch die Anwesenheit von niedrigeren Rhodanokom- plexen in 0,14 molarem Natriumrhodanid erklärt werden, die bei steigender Rhodan- ionenkonzentration immer vollständiger in Tetrarhodanokomplexe umgewandelt werden. Denn aus den Potentialmessungen wissen wir, dass das Aurirhodanid in 0,14 mo- larem Natriumrhodanid schon hauptsächlich als Tetrarhodanokomplex vorhanden ist, und selbst wenn wir annehmen, dass 20 % des Goldes in Form von niedrigeren Kom- plexen vorhanden wären, und dass diese vollständig farblos wären, würden wir höch- stens eine Änderung des Extinktionskoeffizienten von 125 bis 125 -°/, = 156 erhalten, was vollständig ungenügend ist. Wenn wir dagegen annehmen, dass 10 % des Aurirho- danids in 0,14 molarem Natriumrhodanid als stark gefärbte höhere Komplexe vorhan- den sind, ist es leicht, die grosse Steigerung des Extinktionskoeffizienten bei steigender Rhodanionenkonzentration zu verstehen; denn die stark gefärbten rhodanreicheren Komplexe müssen bei steigender Rhodanionenkonzentration immer mehr vorherrschen. Nachweis der Existenz der Penta- und Hexarhodanokomplexe durch Löslichkeitsbestimmungen. Nachdem wir aus unseren Extinktionsmessungen geschlossen hatten, dass Aurirhodanid bei einer Rhodanionenkonzentration von 0,14 m 6* 44 nachweisbare Mengen von Penta- und Hexarhodanokomplexen enthalt, und dass die Menge dieser höheren Komplexe bei einer Rhodanionenkonzentration von 2m recht bedeutend sein muss, haben wir gesucht, dieses Resultat durch andere Messungen zu bestätigen. Durch einige Löslichkeitsbestimmungen von Natriumaurirhodanid in Lösungen mit verschiedenen Rhodanionenkonzentrationen haben wir diesen Zweck erreicht. In zwei Lösungen, die mit Natriumaurirhodanid, NaAuRh,, gesättigt sind, und die dieselbe Natriumionenkonzentration besitzen, muss das Tetrarhodanoauriation auch dieselbe Konzentration besitzen. Hierbei ist doch vorausgesetzt, dass die Aktivitats- koeffizienten der Ionen in den Lösungen dieselben sind; wenn man aber nur dafür sorgt, dass die Lösungen dieselbe Gesamtionenkonzentration besitzen, ist diese Voraussetzung mit guter Annäherung erfüllt. Nun finden wir aber, dass mit steigender Rhodanionen- konzentration die Löslichkeit des Natriumaurirhodanids stark zunimmt, selbst wenn man mit derselben Natriumionen- und bei derselben Gesamtionenkonzentration arbeitet, und man darf daraus schliessen, dass der Tetrarhodanokomplex mit steigender Rhodan- ionenkonzentration in hohem Masse in höhere Komplexe umgewandelt wird. In der Tabelle 27a sind die Resultate einiger solchen Löslichkeitsbestimmungen zusammengestellt. Tabelle 27a. Löslichkeitsbestimmungen von Natriumaurirhodanid. (Aufl) Zusammensetzung der angewandten Mischung nach Rotieren 0,2 m NaRh, 1,8 m NaCl, 0,2 m HCl, 0,00896 m Au”, 0,00448 m Au? 0,00030 04 — 1,6 — 0,2 — 0,00448 — 0,00448 — 0,00038 10 — 1,0 — 0,2 — 0,00448 =— 0,00448 — 0,00075 20 — 0,0 — 0,2 -- 0,00448 — 0,00448 — 0,00178 0,08 m NaRh, 0,32 m NaCl, 0,2 m HCl, 0,00896 m Au?!!, 0,00448 Au? 0,00103 0,2 — 0,2 — 0,2 — 0,0086 — 0,00448 — 0,00118 O4 — 00 — 0,2 — 0,0086 — 0,00448 — 0,00156 Die untersuchten Mischungen wurden aus 4 m Natriumrhodanid, 4 m Natrium- chlorid, 2 m Salzsäure und 0,448 m Wasserstoffaurichlorid dargestellt. Der Auro- rhodanidgehalt der Lösungen wurde durch Zusatz von Wasserstoffaurichlorid und Entfärbung mit gerade der notwendigen Menge Sulfit erhalten. Die Salzsäure und das Aurorhodanid wurden den Lösungen zugefügt, um die Autoreduktion und die Dis- soziation des Aurirhodanids zu vermindern. In allen Mischungen entstand beim Zusatz von Wasserstoffaurichlorid ein Niederschlag von Natriumaurirhodanid. Die Mischungen wurden damit einige Zeit bei 18° rotiert; darauf wurden sie schnell filtriert, und in den Filtraten wurde der Aurirhodanidgehalt sogleich sulfittitriert. Bei den Versuchen wurden Rotationszeiten von 15 Minuten bis 2 Stunden angewandt, ohne dass die Resultate sich mehr als den Versuchsfehlern entsprechend änderten. Die Filtration dauerte einige Minuten, und die Autoreduktion ging in den Filtraten so langsam vor sich, dass Proben, die erst 5 Minuten nach dem Filtrieren titriert wurden, dieselben Resultate wie die sogleich titrierten ergaben. Nach den in Abschnitt VI entwickelten Formeln der Geschwindigkeit der Autoreduktion kann man auch berechnen, dass die 45 Autoreduktion in den angewandten Lösungen genügend langsam verläuft. Die Sulfit- titrierungen sind wegen der kleinen Gehalte der Lösungen an Aurirhodanid ziemlich ungenau, und die Unsicherheit der angeführten Zahlen beläuft sich auf mehrere Ein- heiten in der letzten Dezimale. Aus den Versuchsresultaten kann man annäherungsweise die Konzentrationen der höheren Rhodanokomplexe in den Lösungen berechnen. In Tabelle 27 b sind die Ergebnisse unserer Berechnung der Versuche in 2m Natriumionenlösung mitgeteilt. Die angeführten Rhodanionenkonzentrationen, [Rh ], sind unter Berücksichtigung des ans Gold gebundenen Rhodans berechnet. In den zwei verdünntesten Lösungen (1 und 2) ist wahrscheinlicherweise neben dem Tetrakomplex hauptsächlich Pentakomplex vor- handen. Wenn man annimmt, dass nur Tetra- und Pentakomplexe vorhanden sind, kann man die Konzentration des Tetrakomplexes in diesen Lösungen, [AuRh, ], nach der folgenden Gleichung berechnen: [Au], —[AuRh,] _ [Rh], [Au], —[AuRh, ] TRE E wo [Au] und [Au], die in 1 und 2 gefundenen Aurirhodanidkonzentrationen sind; denn in diesen gesättigten Lösungen ist die Konzentration des Tetrakomplexes überall dieselbe, und die Konzentration des Pentakomplexes ist der Rhodanionenkonzentration proportional. Man findet in dieser Weise für [Au Rh,] 0,000257, abgekürzt 0,00025. Tabelle 27 b. 2 m Na*. Nr. NaRh [Rh | [Aul!!| [AuRh,] [AuRh, ] [AuRh, sr gefunden berechnet 1 02m 0,155 0,00030 0,00025 0,00004 0,00001 0,00030 2 04m 0,377 0,00038 0,00025 0,00009 0,00004 0,00038 3 10m 0,973 0,00075 0,00025 0,00023 0,00026 0,00074 4 20m 1,973 0,00178 0,00025 0,00049 0,00107 0,00181 L ya Au Bh, = 0,0005 ; K su Rh, — 1,0; ‘Kay Ri, = 1,1. Die Differenzen zwischen [Au”/] und 0,00025, d. h. die Gesamtkonzentration der höheren Komplexe, steigen bei den höheren Rhodanionenkonzentrationen viel schneller als die Rhodanionenkonzentration, und diese Lösungen müssen deshalb eine bedeutende Menge Hexakomplex enthalten. Wenn die Komplexitätskonstanten der Penta- und Hexakomplexe K 442, und Kaur, genannt werden, = Cut 1 K PNA RE] NER DR] 0. BAR JL] gilt für alle vier Lösungen: [Auz] — 0,00025 + (1+ K 44 28, [RH] + Kaurm Kaunn, [Rh J’). Aus den [Au”/7]-Werten der 2. und 4. Lösung berechnet man mittels dieser Formel: Kaur, — 1,0; Kum = 1,1. 46 Die mit diesen Konstanten und mit [Au Rh,] —0,00025 für [AuRh, |] und [AuRh, |] berechneten Werte sind in der Tabelle 27 b angeführt. Durch Summation von [AuRh,], [AuRh, |] und [AuRh, ] ergeben sich die ebenfalls angeführten berechneten Werte für [Au/“], die mit den gefundenen Werten dieser Grösse aus- gezeichnet übereinstimmen. Die Löslichkeitsbestimmungen in 0,4 m Natriumionenlösung stimmen auch mit den gefundenen Komplexitätskonstanten gut überein, wie man aus Tabelle 27 c sehen kann, wo die Konzentrationen der höheren Komplexe mit diesen Konstanten berechnet sind. Der hier für [AuRh,] angewandte Wert ist 4mal grösser als der früher angewandte; die Natriumionenkonzentration ist ja aber auch hier 5mal kleiner. Das Löslichkeitsprodukt, Lyaaurn, — [Na']- [AuRh, ], ist bei 2m Na” 0,0005 und bei 0,4 m Na* 0,0004. Der Unterschied ist durch die Verkleinerung der Aktivitätskoeffi- zienten der Ionen mit steigender Ionenkonzentration zu erklären. Tabelle 27 c. 0,4 m Nat Nr. NaRh : [Rh } [Au]. [AuRh;] [AuRk, 7) ‚[AuRk, = [Ari] gefunden berechnet 5 008m 0,035 0,00103 0,00100 0,00003 0,00000 0,00103 6. 02 m 015 0,00118 0,00100 0,00015 0,00003 0,00118 7 04 m 0373 0,00156 0,00100 0,00037 0,00015 0,00152 LxaAurn, — 0,0004 ; K sun, = 1,0; K au rh, = 1,1. Vergleich mit den spektrophotometrischen Messungen, Wenn man die Komplexitätskonstanten kennt, kann man die Bruchteile des Aurirhodanids be- rechnen, die bei einer bestimmten Rhodanionenkonzentration als Tetra-, Penta- und Hexakomplex vorhanden sind. Wir haben diese Bruchteile, Turn Taurs> Taurn,? für die Rhodanionenkonzentrationen berechnet, für welche wir aus Tabelle 27 die molaren Extinktionskoeffizienten des Aurirhodanids kennen. Die Resultate sind in Tabelle 27 d angeführt. Wenn man die Extinktionskoeffizienten der drei Komplexe m Ekyurn» mEkaurn; MEKkaurr, nennt, gilt für den Extinktionskoeffizienten einer Lösung mEk = mEKgurn,* Vaurn + MEK gu rn, * Au Rhy MEK aunty * CAu Rho Die mit den Werten m Ekaurn, = 108, m Ekaurn, —218; mEkaurn, — 248 berechneten Extinktionskoeffizienten sind in Tabelle 27 d angeführt. Sie stimmen in den meisten Fällen mit den beobachteten gut überein. Die bedeutende Abweichung bei der grössten Rhodanionenkonzentration kann in diesem Falle wahrscheinlicherweise durch die Unsicherheiten von sowohl Beobachtung wie Berechnung erklärt werden. Die kleine Abweichung bei der kleinsten Rhodanionenkonzentration kann vielleicht durch eine Hydrolyse erklärt werden. 47 Tabelle 27 d. Die Berechnung des molaren Extinktionskoeffizienten des Aurirhodanids in seiner Abhangigkeit von der Rhodanionenkonzentration. Rh TAuRhy T4uRh, TAuRh, mEk mEk Diff. berechnet beobachtet ber.—beob. 0,010 0,990 0,010 0,000 109 102 + 7 0,026 0,973 0,026 0,001 111 109 + 2 0,038 0,961 0,037 0,002 112 112 0 0,110 0,890 0,098 0,012 126 124 + 2 0,122 0,878 0,107 0,014 127 126 + 1 0,250 0,759 0,190 0,052 141 142 — 1 0,262 0,748 0,196 0,056 142 145 — 3 0,782 0,408 0,319 0,274 186 182 +4 2,0 0,135 0,271 0,595 226 219 + 7 4,0 0,044 0,177 0,778 241 282 41 Kaur, = 1,0; Kur, — 11; mEkgun, = 108; mEkgurn, — 218; mEkyunn. — 248. Die Hydrolyse des Aurirhodanids. Es liegt nahe, daran zu denken, dass das Aurirhodanid nach der Gleichung AuRhz + H,O = AuRh,OH~ + H+ + Rh etwas hydrolysiert ist. Eine Hydrolyse dieser Art ist nach Arbeiten von Hitrorr und von KoHLrAuScH beim Aurichlorid sicher vorhanden’). Wir haben selbst für die Hy- drolysenkonstante des Tetrachloroauriations folgenden Wert gefunden?): P [AuCl,OH ]-[H*]- [CI] _„- =" Ky = — nor ser 0,55-10 7. Eine solche Hydrolyse ist natürlich beim stark komplexen Aurirhodanid weniger zu er- warten als beim schwächer komplexen Aurichlorid; denn das Hydroxylradikal verdrängt schwieriger das stark gebundene Rhodanradikal als das schwächer gebundene Chloratom; eine messbare, wenn auch geringe, Hydrolyse wäre aber doch wohl nicht unwahrschein- lich. Um eine solche Hydrolyse festzustellen, haben wir die Farbe oder vielmehr den Extinktionskoeffizienten des Aurirhodanids in Lösungen mit wechselnder Wasserstofl- ionenkonzentration untersucht. In Tabelle 28 sind die molaren Extinktionskoeffizienten des Aurirhodanids in einigen Lösungen mit demselben Rhodanionengehalt, aber mit wechselndem Wasser- stoffionengehalt zusammengestellt. Die Zahlen gelten für den Darstellungsaugenblick der Lösungen und sind den Tabellen 14, 19, 20 entnommen. 1) HITTORF und Sarkowskı: Zeitschr. physik. Chem. 28, 546 (1899). Konrrausch: ib. 33, 257 1900). *) Vgl. den Anhang dieser Abhandlung. 48 Tabelle 28. Anfangswerte fiir den molaren Extinktionskoeffizienten des Aurichlorids bei wechselndem Sauregehalt. À = 578 pp. Temp. mEk 0,2 m HCl, 0,28 m NaRh, 0,005 m Au’!!! 16° 124 0,5°. = Oe num 22,5° 130 UN READ a 23° 131 0,2 m HCl, 0,28 m NaRh, 0,005 m Au", 0,005 m Au! 16° 135 0,5). = 028 eee, 01005 Mee ee a Omnia 20° 137 1,0. "oe 0005 | AUTRE 22,5° 142 Aus der Tabelle sieht man, dass eine Abnahme des Säuregehaltes von 1m HCl bis 0,2m HCl den Extinktionskoeffizienten nur um ca. 5% ändert. Der Tetrarhodano- auriatkomplex kann daher in diesen Lösungen kaum stark hydrolysiert sein. Wir sind doch geneigt, eine geringe Hydrolyse anzunehmen. Eine solche Hydrolyse kann nicht nur die kleine, aber deutliche Farbenabschwächung bei der Verminderung des Säuregehaltes erklären, sondern kann auch einen Teil der Farbenschwächung erklären, die sich bei kleinen Rhodanionenkonzentrationen in 1 molarer Salzsäure zeigt (vgl. Tabelle 27d), Eine Verminderung der Rhodanionenkonzentration muss ja auf die Hydro- lyse denselben Einfluss ausüben wie eine Verminderung der Wasserstoffionenkonzen- tration. Wir schätzen, dass das Aurirhodanid in 0,1m HCl, 0,1m NaRh noch haupt- sächlich als Tetrarhodanoauriatkomplex vorhanden ist, wenn auch möglicherweise 20% davon hydrolysiert sind. Unsere Resultate über den Zustand des Aurirhodanids in seinen Lösungen sind also die folgenden. In stark sauren Lösungen mit einer Rhodanionenkonzentration von 0,2 m ist das Aurirhodanid hauptsächlich als Tetrarhodanoauriation vorhanden. Wenn man die Rhodanionenkonzentration wesentlich über 0,2 m vergrössert, werden grosse Mengen Penta- und Hexarhodanokomplexe gebildet, und wenn man umgekehrt die Was- serstoffionenkonzentration oder die Rhodanionenkonzentration oder beide verkleinert, macht sich unter Bildung eines Monohydroxotrirhodanoauriations eine Hydrolyse geltend. VII. Die Autoreduktion. Die Oxydationsprodukte des Rhodans. Aurirhodanid ist in Lösung nicht haltbar. Es wird schon bei gewöhnlicher Temperatur recht schnell zu Aurorhodanid reduziert unter gleichzeitiger Oxydation von Rhodan. Wir haben diesen Vorgang die Autoreduktion genannt. Um die Reaktionsgleichung der Autoreduktion zu finden, haben wir folgende Versuche ausgeführt. Festes Kaliumaurirhodanid wurde unter Schütteln in 0,1 m Kaliumrhodanid gelöst und die erhaltene Lösung bis Entfärbung aufbewahrt und analysiert. In einem Ver- 49 such war die erhaltene Lésung 0,00412m Au, die Entfärbung dauerte drei Stunden bei 25°, und die Lösung enthielt für je drei Goldatome 1,03 Moleküle Schwefelsäure und 8,0 Äquivalente Säure. In einem anderen Versuch war die erhaltene Lösung 0,00765 m Au und enthielt für je drei Goldatome 1,00 Moleküle Schwefelsäure und 8,0 Äquivalente Säure. In einem dritten Versuche war die erhaltene Lösung. 0,0162 m Au; hier war ein wenig gelber Niederschlag gebildet, und die Säuremenge betrug 8,1 Äquivalente pro drei Goldatome. In einem Versuch, wo die Autoreduktion des Kalium- aurirhodanids in 1m Kaliumrhodanid vor sich ging, wurden 7,6 Äquivalente Säure pro je drei Goldatome gefunden. Andere Versuche wurden mit Zusatz von Auri- chlorid zu Kaliumrhodanidlösung und nachheriger Entfärbung durch Erwärmung an- gestellt. 10cm? 0,1m Aurichlorid wurden langsam zu 190 cm’ 0,1m Kaliumrhodanid gesetzt. Die ersten Tropfen entfärbten sich sogleich, aber nach und nach wurde die Lösung rot, und zuletzt kam ein roter Niederschlag; bei Erwärmen zu 40° wurde der Niederschlag gelöst und die Lösung entfärbt; in der Lösung war dann pro je drei Goldatome 1,01 Molekül Schwefelsäure gebildet. Als derselbe Versuch mit Im Kalium- rhodanid ausgeführt wurde, betrug die Schwefelsäurebildung 1,00 Molekül pro je drei Goldatome. Die Schwefelsäure wurde als Baryumsulfat gewogen, und die Säure wurde mit 0,1m Natriumhydroxyd und Phenolphtaléin titriert. Bei dieser Titrierung macht die Gegenwart von Aurorhodanid keine Schwierigkeit (vgl. Seite 30). Nach allen unseren Versuchen bildet sich für je drei reduzierte Goldatome recht genau ein Molekül Schwefelsäure. Hiernach dürfen wir annehmen, dass die Oxydation des Rhodanwasserstoffs nach folgender Gleichung vor sich geht: HCNS+30+H,0 = HCN +H,S0, Wenn man mit heissen, konzentrierten Lösungen arbeitet, kann man den entstandenen Cyanwasserstoff riechen. Wenn die Oxydation des Rhodanwasserstoffs nach dieser Gleichung stattfindet, und wenn man berücksichtigt, dass Aurirhodanid als Tetra- rhodanoauriation, AuRh,, und Aurorhodanid als Dirhodanoauroation, Au Rh,, vor- handen ist, erhält man für die Autoreduktion die folgende Gleichung: 3AuRh, + 4H,0 — 3AuRh, + 5HRh+HCN + H,SO,. Nach dieser Gleichung bilden sich fiir je drei reduzierte Goldatome 8 Aquivalente Säure, was unsere Titrierungen denn auch bestätigt haben. Man könnte einen Augenblick geneigt sein, zu meinen, dass Cyanwasserstoff bei unseren Titrierungen nicht mittitriert wurde, da er als Säure so schwach ist, dass er in rein wässriger Lésung mit Phenolphtaléin nicht titriert werden kann. In Gegenwart von Aurorhodanid wird der Cyanwasserstofl indessen in neutraler und basischer Lösung als Aurocyanid gebunden, und daher lässt sich Cyanwasserstoff in Gegenwart von Aurorhodanid glatt titrieren. Unter gewissen Umständen kann die Oxydation des Rhodans durch Aurigold, jeden- falls teilweise, andere Wege nehmen, und zwar namentlich in sehr verdiinnten und in sehr konzentrierten Lösungen. Wenn man z. B. festes Kaliumaurirhodanid in reinem D.K.D. Vidensk. Selsk. Skr., naturvidensk. og mathem. Afd., 8. Række, V. 1. 7 50 Wasser löst, so trübt sich die Lösung nach einiger Zeit, und es scheidet sich in geringer Menge ein weisser, amorpher Körper aus, der beim Stehen dunklere Farben annimmt. Auch in 0,01 molarem Kaliumrhodanid verhält sich Kaliumaurirhodanid ähnlich; dagegen halten seine Lösungen in 0,1 und 1 molarem Kaliumrhodanid sich klar. Wenn die Lösung in bezug auf Rhodanid sehr konzentriert ist, und besonders wenn sie auch stark sauer ist, tritt ein gelber, unlöslicher Niederschlag als Oxydations- produkt des Rhodans auf. Eine Lösung 2m NaRh, 0,2m HCl, 0,001 m HAuCl, wird nach einiger Zeit davon unklar, und die Lösung 4m NaRh, 0,2m HCl, 0,001 m H AuCl, scheidet den gelben Niederschlag schnell aus. Über die Titrierung von Aurirhodanid mit Sulfitlösung. Um die Reaktionsgeschwindig- keit bei der Autoreduktion zu bestimmen, haben wir das Verschwinden des Auri- rhodanids durch Sulfittitrierung verfolgt. Der rote Aurirhodanidkomplex wird durch Zusatz von Sulfit fast augenblicklich zu dem farblosen Aurorhodanidkomplex reduziert, und diese Auroverbindung wird bei gewöhnlicher Temperatur nicht oder nur sehr langsam von einem Sulfitüberschuss weiter reduziert. Die Reduktion geht so glatt und scharf vor sich, dass man mit guter Genauigkeit die Menge von Aurigold in einer sauren, rhodanionenhaltigen Lösung durch Titration mit einer verdünnten Lösung von Natriumsulfit bis zur Entfärbung bestim- men kann. Als Titerflüssigkeit haben wir gewöhnlich eine 0,01 a 0,05 molare Lösung von Natriumsulfit benutzt. Sie wurde von ausgekochtem Wasser dargestellt, unter Kohlen- dioxyd aufbewahrt und mittels Kohlendioxyddruck in die Bürette eingedrückt. Auf diese Weise behandelt, hielt sie ihren Titer unverändert mehrere Stunden hindurch; jeden Tag musste aber der Titer neu bestimmt werden. Die Lösung wurde auf Kalium- jodat eingestellt, dessen Reinheit durch sublimiertes und getrocknetes Jod kontrolliert worden war. Die Titration wurde bei gewöhnlicher Temperatur ausgeführt. Der End- punkt kann mit einiger Übung genau bestimmt werden, wenn man berücksichtigt, dass die Reduktion des Aurigoldes durch Sulfit zwar sehr schnell, aber doch nicht momentan vor sich geht. In der Nähe des Endpunktes muss man die Sulfitlösung tropfenweise zusetzen und nach jedem Zusatz etwas warten, um zu sehen, ob Entfärbung eintritt, zuletzt etwa eine halbe Minute. Je tiefer die Temperatur ist, und je mehr Gold und Rhodan die Lösung enthält, um so langsamer tritt die Entfärbung ein. In Tabelle 29 sind die Resultate einer Reihe Titrierungen gesammelt, durch welche wir uns von der Brauchbarkeit der Methode überzeugt haben. Das Alter der Lösungen ist vom Zusatz des Goldchlorids bis zu Anfang der Titrierung gerechnet. Die Kon- zentration der benutzten Goldchloridlösung wurde durch Fällung des Goldes mit Schwefeldioxyd und Wägung des gefällten Goldes bestimmt. Aus der Tabelle geht hervor, dass alles zugesetzte Aurigold als Aurirhodanid titriert wird. Da nach unseren früher besprochenen Messungen ein Teil des Auri- rhodanids in Aurorhodanid und freies Rhodan dissoziiert ist, so zeigen also diese Ver- suche, dass man durch Sulfittitrierung die Summe des dissoziierten und des nicht dissoziierten Aurirhodanids bestimmt. 51 Tabelle 29. Kontrollanalysen zur Sulfittitrierung. Sulfitlösung Zusammensetzung der titrierten Lösung M : 3 Menge in cm Vol.inem® HAuCl, NaRh HCl Alter Konz. angewandt berechnet 100 0,00199 m 0,4 m lm 3 Min. 0,03412 m 5,80 5,83 50 0,00398 m 0,4 m lm 3 — 0,03412 m 5,80 5,83 50 0,00518 m 0,28 m 1m 0 — 0,05 m 5,15; 517 5,18 50 0,00518 m 0,28 m 0,5 m 0 — 0,05 m 5,18 5,18 50 0,00518 m 0,14 m 1m 0 — 0,05 m 5,20; 5,18; 5,23 5,18 50 0,00518 m 0,056 m 1m 0 — 0,05 m 5,19; 5,20 5,18 50 0,01126 m 0,2 m 0,8 m 1 — 0,01 m 56,0; 56,5 56,3 Gesehwindigkeitsmessungen. Die angewandten Aurirhodanidlösungen wurden aus Salzsäure, Natriumrhodanid, Goldchlorid und Wasser dargestellt. Die Goldchloridlösung war mit Natriumhydroxyd gegen Lakmuspapier neutral gemacht und besass annähernd die Zusammensetzung AuCl(OH), +3NaCl; sie war frisch dargestellt gelb, und wurde beim Stehen dunkelbraun. Das Goldchlorid wurde erst zugesetzt, nachdem die Flasche mit der übrigen Mischung in einem Wasserbad angebracht war, dessen Temperatur durch Zusatz von kaltem oder heissem Wasser auf 18° gehalten wurde. Das Alter der Lösung wurde vom Zusatz des Goldchlorids an gerechnet. Bei den Sulfittitrierungen wurde ca. 0,01 molares Sulfit angewandt. Die titrierten Portionen wurden je für sich hergestellt und nicht aus einer grossen Menge herauspipettiert. Die Tabellen 30—33 enthalten die gewonnenen Resultate. Zuerst ist das Alter der Lösung, {, in Minuten angegeben; « bedeutet den zur Zeit ¢ infolge Sulfittitrierung autoreduzierten Bruchteil des Goldes. Doppelbestimmungen von «a zeigten oft untereinander Abweichungen von 0,01 a 0,02, was wohl hauptsächlich durch ungenügende Temperaturkonstanz bedingt wurde. Der Temperaturkoeffizient der Autoreduktion ist nämlich sehr gross. In 0,004 m Au’, 0,116m NaRh, 0,1m HCl war nach 10 Minuten bei 10° 4 — 0,367, und aus Tabelle 31 berechnet man durch Interpolation, dass derselbe „-Wert bei 18° nach 1,4 Minuten erreicht wird. Hiernach verläuft die Autoreduktion bei 18° 7mal schneller als bei 10°. Tabelle 30. Autoreduktion in 0,004 m Au’!’, 0,316 m NaRh, 0,1 m HCl bei 18°. t a k 1 Min. 0,12 (0,542 - 104) DR Lir 0,155 0,432 — ds 0,185 0,414 — = 0,205 0,389 — (FREE 0,220 0,366 — 10 == 0,295 0,390 — 15 = 0,335 0,383 — EE 0,400 0,337 — 45 — 0,440 0,312 — 60 — 0,480 0,324 — 1200 — 0,565 0,319 — Ir 0,620 0,328 — Mittel: 0,360 - 104 Anm. [Rh] wächst von 0,300 bis 0,308, [H*] von 0,092 bis 0,105. 52 “Tabelle‘ dr. Autoreduktion in 0,004 m Au”, 0,116 m NaRh, 0,1 m HCl bei 18°. teas a k 1 Min. 0,325 (5,21 - 104) 3 — 0,455 5,30 — 5 — 0,50 4,56 — 10 — 0,563 3,77 — 15 — 0,634 4,41 — 30 — 0,721 4,52 — 60 — 0,797 4,52 — 120 — 0,860 4,48 — 180 — 0,884 4,15 — 240 — 0,901 3,89 — 1080 — 0,966 1) (3,47 — ) 2640 — 0,9837 1) (3,44 — ) 4320 — 0,9951 1) (7,15 — ) Mittel: 4,4 . 104 Anm. [Rh ] wächst von 0,100 bis 0,108, [H*] von 0,092 bis 0,103. Tabelle 32. Autoreduktion in 0,002 m Au?!!, 0,308 m NaRh, 0,1 m HCI bei 18°. Te a. k . 3 Min. 0,215 0,434 10° Bj 0,25 0,360 — 10) = 0,32 0,330 — a = 0,375 0,341 — ie = 0,46 0,328 — GO) =. 0,577 0,378 — 100 0,635 0,307 — 10 0,704 0,352 — 2550 — 0,942 1) (0,402 — ) Mittel: 0,354 108 Anm. [Rh ] wächst von 0,300 bis 0,304, [H*] von 0,096 bis 0,101. Tabelle 33. Autoreduktion in 0,002 m An TT 0,108 m NaRh, 0,1 m HCl bei 18°. t a k 1 Min. 0,349 (4,16 - 10%) 3 — 0,471 3,56 — 5 — 0,533 3,56 — 10 — 0,628 3,49 — 15 — 0,686 3,64 — 30 — 0,775 3,89 — 60 — 0,843 3,86 — 120 — 0,892 3,55 — 180 — 0,922 3,81 — Mittel: 3,67 - 10* Anm. [Rh] wächst von 0,100 bis 0,104, ]H*] von 0,096 bis 0,101. 1 Aufbewahrt bei Zimmertemperatur (Mai). | D a 53 Geschwindigkeitstheorie. Wir haben in der ersten Zeit nach Anstellung unserer Ge- schwindigkeitsmessungen eine grosse Arbeit darauf eingesetzt, fiir diese Versuche eine Theorie oder auch nur eine Formel zu erhalten, aber vergebens. Erst nachdem wir weit später entdeckt hatten, dass Aurirhodanid recht stark dissoziiert ist, nahmen wir diese Bestrebungen wieder auf und konnten dann sogleich diese Messungen durch die Annahme erklaren, dass die Geschwindigkeit dem Quadrate der freien Rhodankonzen- tration proportional war. ‘Wenn die Anfangskonzentration des Aurirhodanids (dissoziiert und undissoziiert C genannt wird und « wie friiher den durch Sulfittitrierung gefundenen Autoreduktions- grad bedeutet, so gilt: [AuRh,] = (1—a)C—[Rh,], [AuRk,] = aC+[Rh,], und durch Einsetzen dieser Ausdrücke in die Dissoziationsgleichung des Aurirhodanids erhalt man: (aC + [Rh,]) [Rhy] (1 — a) C — [Rh] RK, — 0,9210 (1) Wenn die Geschwindigkeit der Autoreduktion dem Quadrate der Rhodankonzen- tration proportional ist, muss gelten: Le — k-[Rh,]? oder ar = dl (2) wo k eine Geschwindigkeitskonstante und [Rh,] eine durch (1) gegebene Funktion von a ist. Um (2) zu integrieren, ist ein graphisches Verfahren angewandt. Für jeden der benutzten C-Werte (0,002 und 0,004) ist die durch (1) bestimmte Abhängigkeit zwischen a und 1:[Rh,]? auf Millimeterpapier eingezeichnet, was sich leicht ausführen lässt, in- dem man für eine Reihe passend gewählter [Rh,]-Werte sowohl « nach (2), wie 1: [Rh,]? berechnet und diese Werte als Abszissen und Ordinaten benutzt. Nach den erhaltenen Kurven wurden die Werte von 1:[Rh,]’ für a = 0,005, « = 0,015, a — 0.025, .. u.s.w.. bis « = 0,565 bestimmt, und durch Addition von diesen Werten und Divi- sion durch 100 wurden die Werte des Integrals I, für a — 0,01, a — 0,02, ... u’s.w. bis « — 0,57 erhalten. Wenn man erst eine solche Tabelle für / berechnet hat, ist es leicht, aus den bei den Messungen erhalte- nen a-Werten die Geschwindigkeitskonstante zu berechnen. Denn durch Integration von (2) erhält man: k= Pi & Für a-Werte, grösser als 0,5, sind Ka/Ca und Ka/Ca? kleine Zahlen. Man kann in diesem Falle mit guter Annäherung aus (1) den folgenden Ausdruck für 1: [Rh,]” erhalten: 54 LS ile he (x en ) - 2 [Rh,]? x: Ka 1—a Karte (1 — a)? Wenn man diesen Ausdruck in (2) einführt, erhält man: (x Luca ) de — pat Fe TESTS Ki CS eee woraus man weiter durch Integration erhält: LR 2 ee (ei = (i a) A a.) 1—a, 1—a, 2 &k Kalb) wo a, und a, die zu den Zeiten t, und /, gefundenen Autoreduktionsgrade bezeichnen. Diese angenäherte Formel wurde zur Berechnung der Geschwindigkeitskonstante k be- nutzt für alle Werte von «a grösser als 0,57. Die Prüfung der Theorie. Mit Hilfe dieser zwei Verfahren wurden die in den Ta- bellen 30—33 angegebenen k-Werte berechnet. Beim ersten Anblick könnte man geneigt sein, zu meinen, dass die gefundenen k-Werte mehr als erlaubt variieren. Denn selbst wenn man die k-Werte für {= 1 Minute nicht mitrechnet, weil die Zeitbestimmung recht unsicher sein muss, und gleichfalls von den k-Werten für { grösser als 1000 Mi- nuten absieht, weil die Temperatur bei diesen langdauernden Versuchen nicht auf 18° gehalten wurde, so weisen die übrigen Werte doch bedeutende Variationen auf, und sie zeigen einen gewissen Gang, indem die k-Werte im grossen Ganzen mit der Zeit etwas abnehmen. Die Abweichungen sind indessen nicht grösser, als man sie zu er- warten hat. Schon ein Titrierungsfehler von 0,01 in der Bestimmung des Autoreduk- tionsgrades 4, (und genauer sind die Versuche nicht) ergibt Änderungen von über 10 °/o in k, wie man aus der folgenden Zusammenstellung sehen kann: Änderung von k für einen Fehler von 0,01 in a. Für a = 0,10 0,20 0,30 0,40 0,50 Änderung ink 16% 12/0 10%0 9 %o 8 lo Der Tendenz der k-Werte zum Sinken mit der Zeit ist auch leicht zu erklären. Die Geschwindigkeitskonstante k nimmt nämlich mit steigender Rhodanionen- und Wasserstoffionenkonzentration sehr stark ab, wie bald näher entwickelt werden soll, und da.nun die Gehalte an Rhodanionen und Wasserstoffionen während einer Auto- reduktion nicht konstant sind, sondern beide ansteigen, ist eine gewisse Abnahme von k mit der Zeit gerade zu erwarten. In den Anmerkungen zu den Tabellen sind die Werte der Konzentrationen der Rhodanionen und der Wasserstoffionen beim Anfang und nach dem vollständigen Verlauf der Autoreduktion angegeben. Es wäre vielleicht möglich. wenn auch sehr beschwerlich, diese Abnahme von k zu berücksich- tigen, da wir die Variation von k mit der Rhodanionen- und der Wasserstoflionen- konzentration kennen; wir haben aber davon Abstand genommen; die Sachlage ist nämlich noch komplizierter, weil der bei der Autoreduktion gebildete Cyanwasserstoff 59 auch auf k störend einwirkt. Der Cyanwasserstoff bindet etwas vom Aurogolde. Nach dem Seite 28 besprochenen Gleichgewicht zwischen Aurocyanid und Aurorhodanid muss der als Cyanid vorhandene Bruchteil des Aurogoldes mit wachsender Autoreduktion auch etwas anwachsen; d.h. die aus unseren Messungen berechneten Geschwindigkeits- konstanten werden mit dem Vorschreiten der Autoreduktion etwas grösser werden. Hierdurch wird der Einfluss der steigenden Rhodanionen- und Wasserstoffionenkonzen- trationen teilweise aufgehoben. In Anbetracht dieser Umstände scheint es uns berechtigt, aus dem Material der Tabellen 30—33 zu schliessen, dass die Autoreduktionsgeschwindigkeit dem Quadrate der Rhodankonzentration proportional ist, und weiter daraus zu folgern, dass die Auto- reduktion über das freie Rhodan als Zwischenstufe verläuft. und dass k die Geschwin- digkeitskonstante bei der Zersetzung des freien Rhodans ist. Die Abhängigkeit der Geschwindigkeitskonstante von der Rhodanionen- und der Was- serstoffionenkonzentration. Um genügendes Material zur Diskussion der Abhängigkeit der Geschwindigkeitskonstante k von den Rh-- und H+-Konzentrationen zu erhalten, sind in Tabelle 34 und 35 die Resultate einiger vereinzelten Geschwindigkeitsmessungen in verschiedenen Lösungen zusammengestellt. Die Bedeutung der Bezeichnungen ist dieselbe wie in den Tabellen 30—33, und die Berechnung von k ist wie in diesen Tabellen durchgeführt, Tabelle 34. Autoreduktion bei 18°. t a k 0,004 m Au", 0,316 m NaRh, 08m HCl 10Min. 0,045 135 ee = 0, ee 3550 u. De = OA + 10 = 81S 809 Bent — IT +10 =) 0065 316 = OA u A. =. 10° 000 239 | æn pie 106 | L 10 0,03 83 Mittel: ers SN — À 500% 144 ( 119 30 — 0,105 144 | ri 6a 04 "10 — © OO 132 eo OSG = 2 lO 10 —:. O16 1102 ere OG Ot —— 10: — 70885 9810 OMG = Ol — 10 — 7/01 1644 (10 8945 809 | Mittel: eee) OIG LOL = 1 40 — 0285 885 f 848 Tabelle 35. Autoreduktion bei 18°. - t a k 0,00462 m Au”7, 0,1229 m 1/2 BaRh,, 146m HCl 10Min. 0,07 252 ( fo ei u Be à + ARS 0,4022 — % 10— 0272 3420 | Mittel: 30 — 0,380 3140 56 In Tabelle 36 sind alle k-Werte aus den Tabellen 30—35 zusammengestellt, um zu zeigen, wie die Geschwindigkeitskonstante sich mit der Rhodanionen- und der Wasserstoflionenkonzentration ändert. Die angegebenen Rhodanionen- und Wasserstofl- ionenkonzentrationen sind Mittelwerte zwischen den Anfangs- und den Schlusskonzen- trationen während der Autoreduktion. Tabelle 36. Die Abhängigkeit der Geschwindigkeitskonstante k von [Rh ] und [H*]. 0,098 m H* 0,198 m Ht 0,398 m HF 0.798m H* 1,46 m H* 0,102 m Rh 36700 0,104 — 44000 3550 O105 == 3510 0,109 — 252 0,204 — 8910 ) 809 0,302 — 3540 0,304 — 3600 1102 316 135 0,404 — 1644 239 0,504 — 848 119 0,604 — 132 Anm. Die fett gedruckten Zahlen sind Mittelwerte aus mehreren Bestimmungen. Tabelle 36 zeigt bei näherer Betrachtung, dass die Geschwindigkeitskonstante an- nähernd den Quadraten der Rhodanionen- und der Wasserstoffionenkonzentrationen pro- portional ist. In Tabelle 37 ist eine Übersicht über den Wert von k, = k-[Rh-]?- [H+]? gegeben, um zu zeigen, wie genau diese Gesetzmässigkeit erfüllt ist. Tabelle 37. k, = k- [Rh ]?- [HF]. 0,098m Ht 0,198m Ht 0,398m Ht 0,798m Ht 1,46m Ht 0,102 m Rh 3,67 0,104 — 4,57 6,08 0105 — 6,11 0,109 — 6,40 0,204 — 3,56 5,31 0,302 — 3,04 0,304 — 3,20 3,99 4,62 7,93 0,404 — 2,59 6,17 0,504 — 2,07 f 4,79 0,604 — 7,61 Tabelle 37 zeigt, dass k, = k-[Rh-]?- [H+]? nur mit grober Annäherung konstant ist. In 0,098 m H* sinkt k, von 4 bis 2, wenn [Rh ] von 0,1 bis 0,5 steigt; man muss sich hierbei aber vergegenwärtigen, dass in demselben Intervall k von ca. 40000 bis 848 sinkt. Die Geschwindigkeit ändert sich also etwas schneller als [Rh ]?. In 0,398 m H* ist die Proportionalität zwischen Geschwindigkeitskonstante und [Rh besser erfüllt. Auch die Proportionalität zwischen Geschwindigkeitskonstante und [H*] ~ ist nicht genau erfüllt. In 0,304m Rh wächst k, von 3,20 bis 7,93, wenn [H*] von 57 0,t bis 0,8 steigt; bei der Beurteilung dieser Variation von k, muss man sich aber vergegenwärtigen, dass in demselben Intervall k von 3600 bis 135 variiert. Die Geschwindigkeitskonstante ändert sich also etwas langsamer als [H*]~*. Etwas von der gefundenen Inkonstanz von k-[H*]-[Rh |]? kann natürlich auf der Unsicherheit der experimentellen Bestimmung von k beruhen; es ist zur Erklärung der Inkon- stanz auch von Interesse, sich zu erinnern, dass einige k-Werte aus der Anfangsge- schwindigkeit und andere aus der Schlussgeschwindigkeit der Autoreduktion gefunden wurden; aber diese Fehlerquellen allein genügen doch sicher nicht zur vollständigen Erklärung der Inkonstanz in den k,-Werten. Mit grober Annäherung können wir doch für das ganze untersuchte Gebiet mit k, — 5 rechnen; d.h. wir können für die Ge- schwindigkeitskonstante, k, setzen: k er J Dieser Ausdruck für k ist leicht zu deuten. Wir haben k als die Geschwindigkeits- konstante bei der bimolekularen Zersetzung des freien Rhodans aufzufassen. Wenn wir nun annehmen, dass das freie Rhodan ein wenig nach der folgenden Gleichung hydro- lysiert ist: Rh, + H,0 = H++ Rh + HRhO0, und wenn wir weiter annehmen, dass seine Zersetzung in der Weise vor sich geht, dass zwei Moleküle unterrhodanige Säure, H RhO, miteinander reagieren, erhalten wir fiir k folgenden Ausdruck: aay ER [RA PLAT wo k' die Zersetzungskonstante der unterrhodanigen Säure und K die Hydrolysenkon- stante des freien Rhodans ist. Es gilt namlich: dal zen K. di H*]-{[Rh ]-[HRhO] _ — k.[RRJ — K-[HRROF und | RR.) Da dieser Ausdruck die Geschwindigkeitsmessungen nur mit grober Annäherung wiedergibt, sind eine oder mehrere unserer Annahmen nicht genau erfüllt. Es ist wahr- scheinlicherweise notwendig, zu berücksichtigen, dass das Aurirhodanid bei grossen Rhodanionenkonzentrationen höhere Komplexe bildet. Vielleicht erklärt sich in dieser Weise die Abnahme von k, bei den grösseren Rhodanionenkonzentrationen in 0,1 mo- larer Salzsäure. In 0,4 molarer Salzsäure zeigt sich diese Abnahme nicht; wir nehmen an, dass sie hier von einer Zunahme überdeckt wird, indem das Rhodan in einer Weise zersetzt werden kann, die durch Wasserstoff- und Rhodanionen nicht gehemmt wird. Dieser Zersetzungsweg wird sich nur bei grossen Wasserstoffionen- und Rhodan- ionenkonzentrationen gegenüber dem Weg über unterrhodanige Säure geltend machen können. | Die Geschwindigkeit der Autoreduktion bei sehr kleinen Gehalten an Rhodanionen oder Wasserstoffionen. Um unsere Kenntnis der Autoreduktion zu vervollständigen, haben wir einige Geschwindigkeitsmessungen in 0,00462 m Au?!!, 0,03526 n ‘/: BaRh,, 1,46m HCl, in 0,004m Au’, 0,316m NaRh, 0,01m HCl und in 0,004m Au, 0,316 m D.K.D Vidensk. Selsk. Skr., naturvidensk. og mathem. Afd., 8. Række, V.]. 8 58 NaRh, 0,02m HCl angestellt. Da in diesen Lösungen die Rhodanionenkonzentration, bzw. die Wasserstoffionenkonzentration wegen ihrer Kleinheit sich im Laufe der Autoreduktion sehr bedeutend ändert, haben wir diese Messungen in folgender Weise verwertet. Aus den Geschwindigkeitsmessungen haben wir mit Hilfe der Formel 5 + HAE die Konzentration des in kleiner Menge vorhandenen lons berechnet; da die in dieser Weise bestimmten Werte mit der Zusammensetzung der Lösungen in genügender Übereinstimmung waren, haben wir geschlossen, dass die angewandte Formel für k auch in diesen Lösungen eine brauchbare Annäherung ergibt. Das Zahlenmaterial haben wir in Tabelle 38 und 39 gesammelt. k Tabelle 38. Autoreduktion in 0,00462 m Au!’!, 0,03506 n 1/, BaRh,, 1,46m HCl bei 18°. t a kK. 4.) [Rips ber. atisgk 1 Min. 0,085 3320 9,027 3 — 0,140 2430 0,031 5 — 0,205 3360 0,026 10 — 0,255 2870 0,029 15 — 0,305 3160 0,027 30 — 0,350 2400 0,031 60 — 0,430 2420 0,031 120 — 0,498 2150 0,033 180 — 0,535 LOTUS 0,035 1440 — 0,7001) (900) 2520 — 0,7751) (1000) 3600 — 0,8051) (937) 7200 — 0,8751) (1070) 8880 — 0,9001) (1190) Anm. Nach der Zusammensetzung der Lösung wächst [Rh ] während der Autoreduktion von 0,026 bis 0,034. Tabelle, 39. Autoreduktion in schwach sauren Lösungen bei 18°. t a k (H*]ber.ausk 0,004 m Au!!!, 0,316m NaRh, 0,0)lm HCl 10 Min. 0,86 55800 0,0097 0,004 — 0316-7 — 0,02 — 10 5 0,705 118600 0,021 Anm. Das zu diesen Lösungen angewandte Goldchlorid war lakmusneutralisiert und enthielt also annähernd die Verbindung Au Cl(OH), (vgl. Seite 51). [H*] wächst daher in diesen Lösungen von etwa 0,002, bzw. 0,012 bis etwa 0,013, bzw. 0,023, wenn die Autoreduktion zur Ende verlaufen ist. Die Geschwindigkeit der Autoreduktion nach spektrophotometrischen Messungen. In den Tabellen 14—22 besitzen wir ein grosses Material über die Abnahme der Farbe der Aurirhodanidlösungen mit der Zeit. Das Material ist, da es bei etwas verschiede- nen Temperaturen (15°—23°) gefunden wurde, nicht besonders wertvoll; wir haben 1) Aufbewahrt bei Zimmertemp. (März). 59 es aber doch durchgerechnet, teils um die durch Sulfittitrierung gefundenen Geschwin- digkeitskonstanten zu kontrollieren und teils um die Autoreduktionsgeschwindigkeit in stark sauren, rhodanionenreichen Lösungen kennen zu lernen. Aus der spektrophotome- trisch bestimmten Aurirhodanidmenge ist zuerst der Gehalt an freiem Rhodan mit Hilfe der bekannten Dissoziationskonstante des Aurirhodanids berechnet. Die Summe der Kon- zentrationen des freien Rhodans und des photometrisch bestimmten Aurirhodanids gibt uns das noch nicht autoreduzierte Aurirhodanid; die weitere Berechnung der Geschwin- digkeitskonstanten k und k, ist genau wie friiher durchgefiihrt. Wenn die Messungen bei 18° nicht angestellt sind, haben wir die k-Werte unter der Annahme auf 18° um- gerechnet, dass die Konstante 1,27mal grösser pro Grad wird (vgl. Seite 51; 1,27? — 7). In Tabelle 40 sind einige Resultate zusammengestellt. Tabelle 40. Spektrophotometrische Bestimmungen der Autoreduktionsgeschwindigkeit. CAE) [Ri] X ka 0,028 m NaRh, 1m HCl, 0,002 m Au! 0,005 m Au! 1 0,011 41200 5,00 0,056 — 1 :— 0,006: — 0,006 — 1 0,028 8630 6,77 0,056 — 1. — 0,002 — 0005 — 1 0,04 710 4,34 0056 — 1 — 0008 — 1 0,04 4420 7,08 0,056. — nr 0002 “— 1 0,05 2250 5,61 014. 1 — 0006 — 0008 — 1 0,11 370 4,47 DE, 1 — 002 — (0,005 — 1 0,122 320 4,76 014 — 1 — 0006 — 1 0122 26 337 O14. = io 00020 =— 1 0,133 23 3,965 028 — 02 — 000 — 0,2 0,26 1840 4,95 0,28 — 05 — 000 — 0,5 0,26 140 2,36 > Die berechneten Werte von k,(— k-[Rh f?-[H*}?) liegen für alle die in Tabelle 40 zusammengestellten Lösungen in der Umgebung von 5 und bestätigen somit die durch Sulfittitrierungen gefundenen Werte. In allen diesen Lösungen ist [H*]-[Rh ] kleiner als oder gleich 0,133. Wenn man versucht, für Lösungen, in welchen [H*]-[Rh | grösser als 0,133 ist (wir haben 0,28 m NaRh, 1m HCl und 0,8m NaRh, Im HCl untersucht), die Geschwindigkeitskonstante k, zu berechnen, so variieren die erhalte- nen Werte mit dem Vorschreiten der Autoreduktion sehr stark und sind gewöhnlich weit grösser als 5 (bis 80mal grösser). Die Autoreduktion verläuft in diesen Lösungen deshalb sicher auf einem anderen Reaktionsweg als in den verdünnteren Lösungen. Wenn man annimmt, dass die Reaktion auf dem neuen Weg durch die Anwesenheit von Rhodanionen und Wasserstoffionen positiv katalysiert wird, während sie auf dem gewöhnlichen Weg über unterrhodanige Säure dadurch gehemmt wird, versteht man leicht, dass eine Verdoppelung von [Rh ]- [H*] den Hauptumsatz von dem einen Wege auf den anderen verschieben kann. Unser Material ist indessen zur Bestimmung des neuen Reaktionsweges allzu klein und ungenau, und wir haben keine Bearbeitung da- von versucht. 8* 60 VIII. Die Zersetzungsgeschwindigkeit des freien Rhodans. Wie in der Ubersicht näher entwickelt, haben wir das freie Rhodan als.ein zwischen Brom und Jod liegendes, farbloses und unbeständiges Halogen aufzufassen. Wenn man Brom zu einer salzsauren Lösung von Natriumrhodanid setzt, wird es freigemacht, und durch Messung der Abnahme der Fähigkeit der Lösung zur Jodausscheidung haben wir die Unbeständigkeit des freien Rhodans in wässriger Lösung bestimmt. Die Versuche wurden in folgender Weise angestellt. Eine geeignete Menge Brom- wasser wurde zu verdünnter Salzsäure gesetzt und bei 18° in einem Thermostaten an- gebracht. Wenn die Mischung die Temperatur 18° angenommen hatte, wurde sie schnell in eine konzentrierte, salzsaure, gleichfalls auf 18° gebrachte Lösung von Natrium- rhodanid gegossen. Das Brom darf beim Mischen nicht auf Natriumrhodanid treffen können, ohne dass Salzsäure und Natriumrhodanid in Überschuss vorhanden sind, weil das freigemachte Rhodan sonst allzu schnell hydrolysiert wird. Bei dem angewandten Verfahren wurde trotz aller Vorsicht etwas über die Hälfte des vom Brom freigemachten Rhodans beim Mischen zersetzt. In der Mischung wurde der Gehalt an freiem Rhodan durch Abpipettieren von 50 cm? in 5 cm? 5°/o Kaliumjodid und Titrierung mit Sulfit bestimmt. Die Tabellen 41—44 enthalten das gewonnene Versuchsmaterial. Aus den Titrierungsresultaten ist die Geschwindigkeitskonstante für die Zersetzung des freien Rhodans nach der Formel für eine bimolekulare Reaktion berechnet: $ Hin : (eS: Die gefundenen k-Werte variieren innerhalb jeder Tabelle nur unbedeutend, und zwar unregelmässig um ihr Mittel, woraus hervorgeht, dass die Reaktion wirklich bimoleku- lar ist. Wenn die Rhodanionenkonzentration von 1 bis 0,6 sinkt, also 0,6mal kleiner wird, so wird die Geschwindigkeitskonstante k 19/1146 — 1/060 mal grösser; die Geschwindigkeit ist also dem Quadrate der Rhodanionenkonzentration umgekehrt ‘proportional. Wenn die Wasserstoffionenkonzentration von 1 bis 0,5 sinkt, so wird k 879/114 — 2257 mal grösser; die Geschwindigkeit ändert sich also etwas schneller als das Quadrat der Wasserstoffionenkonzentration. Wenn man aus den gefundenen k-Werten k, =k-[Rh ]°-[H+]? berechnet, erhält man, wie aus Tabelle 45 zu ersehen, Werte in der Umgebung von 5. Tabelle 41. Die Zersetzung des Rhodans in Im HCl, 0,2 m NaRh bei 18°. 40 cm? 5 m HCl + 170 cm? H,O + 5cm? Bromwasser (0,214 m) wurden zu 25 cm? 2m NaRh + 10cm* 5m HCI gesetzt, und die ersten 50 cm? zur Titrierung wurden 0,5 Minuten später (zu { — 0) zu Kaliumjodid gesetzt. t Sulfit (0,01868 m) [Rh,] k 0 Min. 3,50 em’ 0,00138 6 = 1,80 — 0,000709 114 110 1,25 — 492 119 Did, = 0,85 — 335 108 43,5 — 0,45 — 177 113 Mittel: 113,5 l Sulfit (0,01868 m) [Rh,| k 0 Sek 3,87 cms 0,001447 100 — 3,06 — 1144 110 292 — 207 — 0774 124 693° — 1,48 — 0553 113 1417 — 0,78 — 0292 116 Mittel: 115,7 Tabelle 42. Die Zersetzung des Rhodans in Im HCl, 0,12 m NaRh bei 18°. 40 cm? 5m HCl + 180 cm? H,0 + 5 cm? Bromwasser (0,214 m) wurden zu 15 cm? 2m NaRh—. 10cm’ 5m HCI gesetzt, und die ersten 50 cm? zur Titrierung wurden 25 Sekunden später (zu t = zu Kaliumjodid gesetzt. t Sulfit (0,01868 m) [Rhy] k 0 Sek. 3,30 cm? 0,001232 82 — 2,17 — 0811 319 207 — 1,53 — 0571 272 319 — 1,04 — 0388 331 561 — 0,65 — 0243 353 Mittel: 319 Tabelle 43. Die Zersetzung des Rhodans in 0,5 m HCI, 0,2m NaRh bei 18°. 20 cm? 5m HCl + 195 cm? H,O + 5 cm? Bromwasser (0,214 m) wurden zu 25 cm? 2m NaRh 5 cm? 5m HCl gesetzt, und die ersten 50 cm? zur Titrierung wurden 24 Sekunden später zu Kalium- jodid gesetzt. t Sulfit (0,01874 m) [Rh,] k 0 Sek. 3,20 cm? 0,001200 108 — 1,30 — 0487 675 226 — 0,74 — 0278 735 476 — 0,46 — 0172 628 Mittel: 679 Tabelle 44. Die Zerzetzung des Rhodans in 0,4 m HCI, 0,2 m NaRh bei 18°. 20cm? 5m HCI — 200 cm? H,0 + 5 cm? Bromwasser (0,214 m) wurden zu 25 cm? 2m Na Rh gesetzt, und die ersten 50 cm? zur Titrierung wurden 22 Sekunden später (zu {= 0) zu Kaliumrhodanid gesetzt. t Sulfit (0,01845 m) [Rh] k 0 Sek 3,13 cm® 0,001154 96 — 1,00 — 0369 1150 i 0,90 — 0332 1080 199 — 0,52 — 0192 1310 Mittel: 1180 Tabelle 45. k ko lm HCl, 0,12 m NaRh 319 4,59 a Fen mee 114,6 4,58 ÉD qe | 679 6,79 he = ce |, = 1180 7,55 62 Alle diese Gesetzmässigkeiten stimmen mit unseren Resultaten aus den Auto- reduktionsuntersuchungen gut überein und bilden somit einen Beweis fiir unsere An- nahme, dass die Autoreduktion über das freie Rhodan verläuft und durch seine Un- bestandigkeit hervorgerufen wird. In Tabelle 46 sind die Resultate einiger Messungen über die Zerzetzung des Rho- dans bei 13,8° aufgeführt. Tabelle 46. Die Zersetzung des Rhodans in 1m HCl, 0,08m NaRh bei 13,8°. 40cm? 5m HCl + 185cm? H,0 + 5 cm? Bromwasser (0,214 m) wurden zu 10cm* 2m NaRh + 10cm? 5m HCl gesetzt, und die ersten 50 cm” zur Titrierung wurden 28 Sekunden später zu Kalium- rhodanid gesetzt. t Sulfit (0,01868 m) [Rh,] k 0 Sek. 3,60 cm? 0,001348 110 — 1,80 — 0672 408 252 — 1,03 — 0384 442 374 — 0,71 — 0265 503 Mittel: 451 Als Mittel erhält man aus dieser Tabelle für k 451 und fiir k, 451 - 0,08? — 2,99. Da nach Tabelle 45 k, in 1m Salzsåure bei 18? 4,59 betrågt, ergibt sich hiernach fir eine Temperatursteigerung von 1” eine Vergråsserung von k, im Verhåltnis 4,2 1 1,50 V 399 bl. Da wir für die Geschwindigkeitskonstante der Autoreduktion pro Grad eine Vergrösse- rung im Verhältnis V = — 1,28 gefunden haben (vgl. Seite 51), also eine weit grössere Zahl, so muss die Dissoziation des Aurirhodanids in Rhodan und Aurorhodanid mit der Temperatur steigen, und zwar muss die Dissoziationskonstante pro Grad etwa 8°/o grösser werden. — 1508. 125 | 1,11 Diese Zunahme ist recht plausibel. Ihr entspricht nach dem zweiten Warmesatz ein Warmeverbrauch bei der Dissoziation des Aurirhodanids von: dink Hs 2.291”. 0,08 — ca. 10000 Kalorien. Q = RT? 63 IX. Die Goldausscheidung. Die chemische Gleichung des Vorgangs. Wenn man eine Goldrhodanidlösung längere Zeit aufbewahrt, so wird das Gold nach und nach in freier Form ausgeschieden. Dieser Vorgang findet hauptsächlich in der Weise statt, dass der Gehalt der Lösung an Auro- rliiodanid in Gold und Aurirhodanid gespalten wird: 3AuRh, = 2Au+ AuRh, +2Rh, wonach das gebildete Aurirhodanid zu Aurorhodanid autoreduziert wird; dieses wird dann wieder unter Goldausscheidung gespalten u.s.w. Den Beweis dieser Reaktions- folge erblicken wir darin, dass Aurorhodanidlösungen, die kein Aurirhodanid enthalten, beim Stehen gleichzeitig Gold ausscheiden und aurirhodanidhaltig werden, und wenn der Gehalt an Wasserstoffionen und Rhodanionen einigermassen gross und die Geschwin- digkeit der Autoreduktion daher klein ist, so wird in Ubereinstimmung mit obiger Gleichung im Anfange des Vorgangs fiir je zwei ausgeschiedene Goldatome ungefahr ein Molekül Aurirhodanid in der Lésung gefunden. Einige Analysenresultate, die dies zeigen, sind in Tabelle 47 enthalten. Tabelle 47. Goldausscheidung und Aurirhodanidbildung in Aurorhodanidlösungen. [Rh] = 0,4; [Ht] = ca. 0,1; T — ca. 14°. Nach 24 Stunden. Nach 48 Stunden. ld ausge- - ee ne de Le, aus 25cm. beob. ber. aus 25cm}. beob. ber. 0,0121 m Aul, 0,425 m NaRh, 0,1 m HCl 1,6 mg 0,00012 0,00016 1,7 mg 0,00013 0,00017 0,0242 — 0,45 — 0,1 — 44 — 0,00036 0,00045 5,7 — 0,00052 0,00058 0,0484 — 0,5 — 0,1 — 12,5 — 0,00100 0,00127 19,8 — 0,00176 0,00201 0,0969 — 0,6 — 01 — Niederschlag aus NaAuRh, Niederschlag aus Na AuRh, Die angewandten Lösungen wurden durch Zusatz von neutralisiertem Goldchlorid (H Au CI, + ca. 3 NaOH) zu Salzsäure und Natriumrhodanid und nachherige Reduk- tion mit Sulfit dargestellt und in Portionen von 25 cm? verteilt. Das ausgeschiedene Gold wurde abfiltriert, geglüht und gewogen, und im Filtrate wurde das Aurirhodanid sulfittitriert. In der Tabelle sind ausser den beobachteten Aurirhodanidkonzentratio- nen, [Au’], auch diejenigen aufgeführt, die man aus den ausgeschiedenen Goldmengen berechnet, unter der Annahme, dass pro zwei ausgeschiedene Goldatome ein Molekül Aurirhodanid gebildet wird. Die beobachteten Aurirhodanidkonzentrationen sind über- all ein wenig kleiner als die berechneten, was aber auch zu erwarten ist, da die Au- toreduktion durch die Anwesenheit der Wasserstoffionen, der Rhodanionen und des Aurorhodanids sicher nicht vollständig gehemmt wird. Das Gleichgewicht Aurorhodanid-Aurirhodanid an einer Goldoberfläche. Wenn man in Tabelle 47 die Resultate nach 24stündigem Stehen mit denjenigen nach 48stündigem Stehen vergleicht, zeigt sich Folgendes. In der goldreichsten Lösung geht die Zersetzung des Aurorhodanids den zweiten Tag fast ebenso schnell wie am ersten Tage; je weniger 64 Gold die Lösung aber enthält, um so mehr nähern sich die für einen und für zwei Tage gefundenen Resultate einander. In der 0,0121 molaren Goldlösung scheint die Reaktion schon nach einem Tage annähernd fertig zu sein: hiernach darf man schlies- sen, dass in dieser Lösung nach zwei Tagen ein Gleichgewicht zwischen Aurorhodanid, Aurirhodanid und Gold fast erreicht wird. Nach der chemischen Gleichung des Vor- ganges: 3AuRh, = AuRh, +2Rh +2 Au kann man folgern, dass die Gleichgewichtsbedingung die folgende Form besitzen muss: [Au Rh, ] [Rh]? les PS [Au Rh, ]? Wenn man annimmt, dass in der verdünntesten Lösung das Gleichgewicht nach zwei Tagen vollständig erreicht worden ist, berechnet man für die Gleichgewichtskon- stante K den Zahlenwert 14,3. 0,00013 - 0,4? RTC — 14,3. Dass in den stärkeren Goldlösungen das Gleichgewicht nicht so schnell erreicht wird wie in der verdünnten, ist leicht zu verstehen, da die zum Gleichgewicht not- wendige Aurirhodanidmenge mit der dritten Potenz der Aurorhodanidkonzentration anwächst. Den Zahlenwert der Gleichgewichtskonstante, K, kann man auch aus Potential- messungen, die wir früher angestellt haben, berechnen. In einer Lösung von Aurorho- danid und Aurirhodanid, die mit Gold in Gleichgewicht ist, müssen die zwei Elek- trodenpotentiale, die den Übergängen Aurirhodanid-Aurorhodanid und Aurorhodanid- Gold entsprechen, einander gleich sein. !) Für das Aurirhodanid-Aurorhodanid-Potential gilt: Due [Au Rk] in AuRh.] [RR]? [Au Rh, ] uRh,]- [Rh J?’ E = En +" = 0,645 +- 0,029 Isa ‘und für das Aurorhodanid-Gold-Potential gilt: In Au Rhy ] eh [AuRh, | RL 0,689 -+ 0,058 lg [Rh F E = oEn et Das Gleichsetzen dieser zwei Potentiale ergibt die gesuchte Gleichgewichts- bedingung: 0,044 ARRETE SE RS [AuRh, ]> Der aus den Potentialmessungen berechnete Zahlenwert 33 ist mehr als doppelt so gross wie der analytisch bestimmte. Bei der Beurteilung der fehlenden Überein- stimmung muss man sich erinnern, dass beide Werte recht unsicher sind. Namentlich 1 Vgl. Lurner: Zeitschr. physik. Chem. 36, 385 (1901). 65 die analytische Bestimmung ist wahrscheinlicherweise sehr ungenau. Das Gleichgewicht hat sich selbst in der verdiinnten 0,0121 molaren Lösung nach zwei Tagen sicher nicht vollstandig eingestellt, und die Bestimmung der Aurirhodanidkonzentration ist wegen der Kleinheit dieser Konzentration ungenau. Wir haben daher den Wert 14,3 ausser Betracht gelassen und mit 33 gerechnet; aber auch dieser Wert ist sehr unsicher, in- dem ein Fehler von nur zwei Millivolt in einer der zu seiner Berechnung angewandten Normalpotentialen den Wert um 17 °/o ändert. Wenn in einer Lösung die Hauptmenge des Goldes als Aurirhodanid, bzw. als Aurorhodanid vorhanden ist, sind die folgenden Formen der Gleichgewichtsbedingung zur Orientierung gut geeignet: [AuRh, P [AuRh,] _ [AuRh, | Tarp» 92% Lau] > °3° [RP (am) — 3" Zu allgemeiner Orientierung ist es das beste, die folgenden Variabeln einzuführen: z=[Rh |: [Au]. und y = [AuRh,]: [Au], wo [Au] die gesamte Goldkonzentration bedeutet. Die Gleichgewichtsbedingung nimmt mit ihnen die folgende Form an: xv? — 33(1—y)*: y. Diese Gleichung ist zum Beispiel durch folgende Wertpaare befriedigt: y Or? 0,1 0,2 0,5 0,8 0,9 0,95 0,99 log x 1,753 1,190 0,913 0,458 —0,241 —0,718 —1,181 —2,239. Mit Hilfe dieser Werte ist die Kurve in Figur 2 (Seite 14) gezeichnet, die eine gute Übersicht über die Änderung des Verhältnisses zwischen Aurirhodanid und Aurorhodanid mit dem Verhältnis zwischen der Rhodanionenkonzentration und der gesamten Goldkonzentration abgibt. Die Geschwindigkeit der Goldausscheidung. Es ist nun gezeigt worden, dass die Goldausscheidung nach der folgenden Gleichung vor sich geht: 3AuRh, = AuRh, + 2Rh + 2Au, und dass dieser Vorgang reversibel ist und zu einem Gleichgewicht führt, für welches gilt: [AuRh,]-[Rh F [Au Rh, F = ca. 33. Um auch die Geschwindigkeit der Goldausscheidung kennen zu lernen, haben wir Versuche bei 40° angestellt. Es wurde vorgezogen, die Versuche bei 40° auszufiihren, da die Reaktion bei 18° sehr langsam verläuft. Zu den Versuchen wurde vorzugsweise eine 0,1 molare Aurorhodanidlésung benutzt, in welcher die Rhodanionenkonzentration 0,3 molar und die Wasserstoffionenkonzen- tration 0,1 molar war; ausserdem wurden aber einzelne Versuche mit anderen Lösungen D. K. D. Vidensk. Selsk. Skr., naturvidensk, og mathem. Afd. 8. Række, V.1. 9 66 angestellt. Die Lösungen wurden aus abgemessenen Mengen Salzsäure und Natrium rhodanid durch Zusatz von neutralisiertem Aurichlorid und nachherige Entfärbung mit Natriumsulfit dargestellt. Das Aurichlorid wurde in kleinen Portionen zugesetzt, die je für sich entfärbt wurden. Nach Verdünnen mit Wasser zu dem gewünschten Volumen wurde die Lösung in Portionen von 50 cm? bei 40° in einem Thermostaten angebracht. Das ausgeschiedene Gold wurde abfiltriert, geglüht und gewogen, und das in der Lösung vorhandene Aurirhodanid wurde durch Sulfittitrierung bestimmt. Tabelle 48 zeigt den Verlauf der Goldausscheidung in 0,1m Au”, 0,5 m NaRh, 0,1m HCl; in dieser Lösung ist die Rhodanionenkonzentration 0,3 m und die Wasserstoflionenkonzentration ca. 0,1 m.!) Während die ausgeschiedene Goldmenge mit der Zeit ständig anwächst, passiert die vorhandene Aurirhodanidmenge durch ein Maximum. Vor dem Maximum wird das Aurirhodanid durch die Goldausscheidung schneller gebildet, als es durch die Autoreduktion zersetzt wird, während die Sachlage nach dem Maximum eine um- gekehrte ist. Tabelle 48. Die Goldausscheidung in 0,1m Au/, 0,5m NaRh, 0,1m HCL. [Rh] —— 0S) [at] = ca. 0,1; t = 40°. Gold aus 50 cm [Au ausgeschieden Nach 2stiindigem Stehen 0,00088 m 24,3 mg — 4 = 0,00176 - 44,0 - — 6 — 0,00222 - 56,6 - — 18 — 0,00308 - 184,2 - 0,00288 - 280,3 - — 42 — 0,00262 - 249,9 - DE den 293 - - ' { 0,00196 - 382 - — 66 — 0,00169 - 371 - | 0,00169 - 413 - Wir haben nicht versucht, eine Formel für die Geschwindigkeit der Goldaus- scheidung aufzustellen. Die Goldausscheidung besitzt nämlich wie die meisten anderen heterogenen Reaktionen einen schlecht definierten Verlauf und gibt bei Wiederholungen nicht genau übereinstimmende Resultate (vgl. Tabelle 48). . Der katalytische Einfluss des freien Goldes. Wir haben gefunden, dass die Goldausscheidung durch die Anwesenheit von freiem Gold beschleunigt wird. Aus einer frischen Lösung, 0,1m Au’, 0,5m NaRh, 0,1m HCl, wurden gleichzeitig zwei Portionen von 50 cm? entnommen und bei 40° angebracht. Zu der einen Portion wurde das aus einer ähnlichen Portion in einigen Tagen ausgeschiedene Gold gesetzt. In dieser Portion stieg die Aurirhodanidkonzentration in zwei Stunden zu 0,00312, während sie in der anderen goldfreien nur zu 0,00108 stieg. Die Oberfläche des Goldes wirkt also als Kontaktsubstanz; es ist daher verständlich, dass die Geschwindigkeit der 1) Dieser Wert ist doch wahrscheinlicherweise etwas zu hoch; das angewandte Aurichlorid war gegen Lakmuspapier neutralisiert, und zu seiner Neutralisation verbraucht Aurichlorid nicht genau 2NaOH pro AuCl,, sondern etwas mehr, 67 Goldausscheidung variieren kann, je nachdem das Gold in einem Versuch mehr oder weniger kompakt ausgeschieden wird. Am schnellsten wiirde die Ausscheidung wohl stattfinden, wenn das Gold sich in kolloider Form einige Zeitlang aufgeschlemmt hielt. Die Abhangigkeit der Goldausscheidung von der Zusammensetzung der Lésung. Wir haben untersucht, wie die Geschwindigkeit der Goldausscheidung sich mit der Konzentration des Aurorhodanids, der Rhodanionen und der Wasserstofl- ionen ändert. Die Resultate sind in den Tabellen 49—51 gesammelt. Tabelle 49. Der Einfluss der Aurorhodanidkonzentration. [Rh] = 0,3; [H*] = ca.0,1; t = 40°. Nach 43,5 Stunden bei 40°, [Au] Gold aus 50 cm? ausgeschieden 0,1m Au’, 0,5m NaRh, 0,1 m HCl 0,00272 m ") 280,8 mg ') 0,05 — ee Ot 0,00090 - 10/58 ee 0,02 — 0,34 — or. — 0,00026 - 47,4 — Tabelle 50. Der Einfluss der Wasserstoffionenkonzentration. [Rh] = 0,3; [Auf] 0,1; t = 40°. Nach 18 Stunden bei 40°. [Au!!] Gold aus 50 cm? ansgeschieden 0,1 m Au‘, 0,5m NaRh, O,lm HCl 0,00308 m 184,2 mg 01 — Dave EL 0,00653 - 2198 — oo = 05: — os — 0,00356 - AT — Tabelle 51. Der Einfluss der Rhodanionenkonzentration. [Ht] = ca. 0,1; [Au] = 0,1; t = 40°. Nach 18 Stunden bei 40°. [Rh] [Au!!7] Gold aus 50 cm? ausgeschieden f 0,00288 270,0 mg 0,1 m Au’, 0,4 m NaRh, 0,1 m HCl 0,2 \.0,00284 308,5 O1 — 0,5 — O1 — 0,3 0,00308 184,2 O1 — 10 — 0,1 — 0,8 0,00162 81,5 Die Tabellen zeigen, dass die Geschwindigkeit der Goldausscheidung mit steigender Rhodanionenkonzentration sinkt. Dagegen ist sie von der Wasserstoffionenkonzentration ziemlich unabhängig und steigt nur ein wenig mit abnehmendem Säuregrad. Wenn man die Wasserstoffionenkonzentration durch Zusatz von Natriumazetat sehr stark herunterdriickt, geht die Goldausscheidung doch, wie wir durch einige qualitative Versuche gefunden haben, mehrmal schneller als in den salzsauren Lösungen. Die in den Lösungen nachweisbare Aurirhodanidmenge nimmt mit der Konzentra- tion der Lösungen sehr stark ab. Wenn man die in den Tabellen 49—51 untersuchten 1) Interpoliert aus Tabelle 48. g* 68 Lösungen 10mal mit Wasser verdünnt, nehmen sie beim Stehen die Farbe des Auri- rhodanids gar nicht an; es werden also keine sichtbaren Mengen von Aurirhodanid gebildet. Eine spätere Versuchsreihe. Einige Monate nach den besprochenen Ver- suchen wurde eine neue Reihe Versuche mit 0,1 m Au’, 0,5 m NaRh, 0,1 m HCl angestellt. Tabelle 52 enthalt die dabei gewonnenen Resultate. In dieser Versuchsreihe ging die Goldausscheidung bedeutend schneller als früher (zu Anfang sogar 4mal schneller). Die gleichzeitig angestellten Parallelversuche stimmten recht gut überein und wiesen jedenfalls keine Variationen auf, die mit dem Unterschied der beiden Reihen vergleichbar waren. Es ist möglich, dass die Salzsäure bei der Darstellung der Lösungen der letzten Reihe erst nach dem Zusatz und der Reduktion des Aurichlorids zugesetzt wurde, und dass die Reduktion zu Aurorhodanid also in neutraler Natriumrhodanid- lösung vorgenommen wurde. Bei der Ausführung der Versuche wurde nicht daran gedacht, dass eine solche kleine Abweichung von Bedeutung sein könnte. Erst viel später sind wir auf den Gedanken gekommen, dass vielleicht ein bei der Reduktion in neu- traler Lösung entstandenes Nebenprodukt die Goldausscheidung beschleunigt hat. Tabelle 52. Die Goldausscheidung in 0,1 m Au’, 0,5 m NaRh, 0,1 m HCl. [Rh] = 0,3; [HF] = ca. 0,1; t — 40°. (Auf Gold aus 50 cm? ausgeschieden Nach 1stündigem Stehen 0,0035; 0,00218 ie mg; 2 — 2 — — 0,00422; 0,00428; 0,00436 96,5 — «— 3 — — 0,00442; 0,00442 105,4 — ; 109 mg — 6 — — 0,00592; 0,00608 146 — — 9 — — 0,00728 ræs — 12 — — 0,00780 239 — — 15 — — 0,00762 — 18 — — 0,00704 Kt — 2 — — 0,00624 340 — — 24 — — 0,00640; 0,00650 260 — — 46 — — 0,00344 352 —; 355 — — al — — 0,00306; 0,00282 384 —; 377 — — 72 — — 0,00166 500 — — 75 — — 0,00150; 0,00150 ee — 92 — URNE SE Ma SURE 486 —; 487 — — 94 — — 0,000934 516 —. — 95 — — 0,00148; 0,00132 478 — = 770 — — 0,00018; 0,00018 it Das Verhalten der Aurirhodanidlösungen. Eine frische Lösung von Aurirhodanid kann Gold auflösen. Wenn man die Lösung stehen lässt, wird das Aurirhodanid nach und nach autoreduziert, und nach einiger Zeit wird ein Zustand erreicht, wo die Lösung mit metallischem Gold in Gleichgewicht sein wird, wo sie also sozusagen mit Gold gesättigt ist. Wenn die Autoreduktion weiter vorschreitet, fängt die Lösung an, Gold auszuscheiden, und es wird nach einiger Zeit ein in bezug auf Aurirhodanid annähernd stationärer Zustand erreicht, wo die durch die Autoreduktion verbrauchte Menge von 69 Aurirhodanid fast vollständig durch die Zersetzung des Aurorhodanids wiedergebildet wird. Ein ähnlicher annähernd stationärer Zustand wird erreicht, wenn man von Auro- rhodanid ausgeht. Mit der Zeit müssen diese beiden Zustände sich asymptotisch an- einander nähern, und der Goldgehalt wird in beiden Lösungen in derselben Weise gegen Null abnehmen. Wir haben diese Schlüsse durch die in Tabelle 53 und 54 zusammengestellten Ver- suche kontrolliert. Aus Tabelle 53 sieht man, wie die Aurirhodanidkonzentration in einer Aurorhodanidlösung sich ändert, in welcher die Aurirhodanidkonzentration an- fänglich etwas grösser war als dem stationären Zustande entsprechend. Tabelle 53. Tabelle 54. 0,0066 m Au 7, 0,0934 m Au, 0,5mNaRh, 0,1mHCI bei 40° 0,1 m Au!, 0,5 m NaRh, 0,1 m HCI bei 40°. Nach Ostündigem Stehen 0,0066 m Au!!! Naeh 2stündigem Stehen 0,0009 m Au!!! Sur — Se -— RAGE RE 2.0.0007, — 12 — — 0,0031 — — 18 — — 0,0031 = Ba — 5,000 ee NT Got Peps 2 = ON — 266 ok AM OS p= Wenn man von derselben, aber nur aurirhodanidfreien Lösung, ausgeht, erhält die Lösung beim Stehen die in Tabelle 54 angeführten Aurirhodanidmengen. ‘Die anfänglich aurirhodanidhaltige und die anfänglich aurirhodanidfreie Lösung enthalten also schon nach etwa 10 Stunden annähernd dieselbe Menge Aurirhodanid. Anhang über das Aurichlorid. Über den Zustand des Goldes in Aurichloridlösungen liegen in der Literatur nur qualitative Angaben vor. Um die Komplexität des Aurirhodanids mit derjenigen des . Aurichlorids vergleichen zu können, haben wir deshalb im Laufe unserer Unter- suchungen über die Rhodanide des Goldes das Normalpotential Gold-Aurichlorid gemessen. Die zu diesem Zwecke angestellten Potentialmessungen haben uns nebenbei erlaubt, die Hydrolyse des Aurichlorids zahlenmässig festzulegen. Die bei unseren Aurichloriduntersuchungen gefundenen Daten sind in diesem kleinen Anhang zusam- mengestellt. Das Normalpotential Aurichlorid-Gold. Das Potential von Goldelektroden in auri- chloridhaltigen Lösungen ist von mehreren Beobachtern untersucht worden (NEUMANN "), Fawsırr?)). Diese älteren Messungen sind indessen ziemlich wertlos; in einigen ist die Zusammensetzung der Goldlösung gar nicht angegeben, und wo die Zusammensetzung 1) Zeitschr. physik. Chem. 14, 193 (1894). 2) Journ. Soc. Chem. Ind. 25, 1133 (1906). 70 angegeben ist, kennt man doch nicht den Zustand des Goldes und den Gehalt an Chlorionen. | Wir. haben daher einige neue Messungen ausgeführt. Bei diesen wurden Lösungen von Wasserstoffaurichlorid benutzt, in welchen die Hydrolyse des Tetrachloroauriat- komplexes durch Zusatz von Salzsäure so stark zurückgedrängt war, dass man die Konzentration des Tetrachloroauriat-Ions gleich der Goldkonzentration und die Chlor- ionenkonzentration gleich der Konzentration der zugesetzten Salzsäure setzen konnte. Tabelle 55 enthält die gewonnenen Resultate. Tabelle 55. Potential von Ketten: Au | Aurichloridlös. | gesatt. KCl | 0,1 m KCl, HgCl| Hg. Bei ca. 17°. Aurichloridlösung E Dif. pot. oE I II III I II III 0,1 m HAuCl,, 1 m HCl 0,627 ate. ee —0,017 0,673 en aR 0,1 — 05 — 0,662 0,647 0,657 —0,012 0,670 0,655 0,665 0,1 — 0,25 — ee 0,672 0,683 —0,009 ms: 0,654 0,665 0,1 — 0,1 — Fee 0,713 0,717 —0,007 SER ‚0,662 0,666 0,1 — 0 — 0,813 sts Wer — 0,004 LE ER ee: 0,05 — 0,5 — a 0,636 pie 2 —0,012 0,650 Mittel: 0,671 0,655 0,665 Für E, das gemessene Potential, sind unter I, II, III die Resultate von drei ver- schiedenen Messungsserien angegeben, die zu verschiedenen Zeiten und mit verschie- denen Goldelektroden ausgeführt wurden. Als Goldelektroden wurden galvanisch ver- goldete, in Glas eingeschmolzene Platindrähte angewandt. Es wurden in .derselben Messungsreihe nur Elektroden benutzt, die in einer Aurichloridlösung Potentiale er- gaben, die auf weniger als 1 Millivolt übereinstimmten. Man sieht sogleich aus der Tabelle, dass die drei Reihen untereinander nicht gut übereinstimmen, indem Abweichungen bis auf 15 Millivolt vorhanden sind. Die gemes- senen Potentiale waren im grossen ganzen auch recht schlecht definiert; sie änderten sich in den ersten Stunden nach der Zusammenstellung der Kette oft viele Millivolt; die eingeführten Werte sind die nach mehrstündigem Stehen gefundenen. Bei einer erneuten Untersuchung dieser Potentiale wird es sicher zweckmässig sein, das Auro- Auri-Gleichgewicht in einer Goldchloridlösung in Berührung mit metallischem Gold genau zu studieren; vielleicht sind in den angewandten Lösungen die Auro-Mengen beim Gleichgewicht so gross, dass die Unsicherheiten der Messungen davon herrühren. Um aus den gemessenen Potentialen ein Normalpotential zu berechnen, haben wir zuerst die Messungen für das Diffusionspotential korrigiert. Der Wert des Diffusions- potentials zwischen der salzsauren Goldlösung und dem gesättigten Kaliumchlorid wurde nach HENDERSON") berechnet, und die gefundenen Werte sind in der Tabelle ange- führt. Aus dem korrigierten Potentialwert, E’, wurde das Normalpotential, „E, nach der Formel 1) Vgl. diese Abhandlung S. 27. 71 [Au CI, ] E = E+ in [ct]! (1) berechnet. Diese Formel beruht auf der Annahme, dass die folgende Reaktion poten- tialbestimmend ist: | AuCI, +38 = Au +4CT. Die nach der Formel (1) berechneten Normalpotentiale sind in Tabelle 55 ange- führt. Innerhalb derselben Messungsreihe sind die für das Normalpotential gefun- denen Werte von der angewandten Goldkonzentration und der angewandten Salzsäure- konzentration unabhängig. Wir dürfen hieraus auf die Richtigkeit der Formel (1) und weiter rückwärts auf die Richtigkeit der angenommenen ee Reaktion schliessen. Die in den drei Messungsreihen gefundenen Mittelwerte des Normalpotentials sind 0,671, 0,655, 0,665. Die angewandten Goldelektroden haben also bis 16 Millivolt ver- schiedene Potentiale ergeben. Das Mittel aus allen Messungen ergibt fiir das Normal- potential 0,664, und wenn dieser Wert durch Addition von 0,337 auf die Normal- Wasserstoffelektrode umgerechnet wird, erhalt man fiir die Reaktion: AuCcl, +36 = Au+4Cl das Normalpotential, „Er, gleich 1,001. Die Hydrolyse des Tetrachloroauriatkomplexes. Wir haben auch das Potential einer Goldelektrode in 0,1 molarem Wasserstoffaurichlorid ohne zugesetzte Salzsäure gemessen. Das Resultat ist in Tabelle 55 zu finden. Aus dieser Messung kann man die Hydro- lysenkonstante des Tetrachloroauriatkomplexes berechnen. Nach Untersuchungen von HittorF und SALKOWSKY und von KOHLRAUSCH muss man annehmen, dass dieser Komplex nach der Gleichung Auch, CH,O = AuCLOH Hr er etwas hydrolysiert ist. Wenn der hydrolysierte Bruchteil x genannt wird, so wird die Chlorionenkonzentration der Lösung gleich 0,1-x und die Konzentration des Tetra- chloroauriations gleich 0,1(1—x). Wir erhalten deshalb zur Bestimmung von x: x 4 0,1(1—7x) Pr CIRE Hier ist E’ das gemessene und für das Diffusionspotential korrigierte Potential; es ist gleich 0,817. „E ist das Normalpotential gegenüber der Dezinormal-Kalomelelektrode; es ist gleich 0,671 gesetzt, indem wir den Mittelwert angewandt haben, der in der- selben Reihe gefunden wurde. Diese Zahlen ergeben: ate 0,817 — 0,671 01-2) _ 10 om — 107; x 0,069. Eine 0,1 molare Lésung von Wasserstoffaurichlorid ist also nach der Gleichung 72 AuCl, HOS AuCLON + Aes 6,9 %0 hydrolysiert. Daraus erhält man für die Hydrolysenkonstante, K;: . [H*]-[ClU]-[AuCl,OH | 0,1069-0,006® er Rz [Au Cl] 4 0,0931 = Oe US Wenn man mit dem gefundenen Wert, K, = 0,55 - 10 4, die Hydrolyse in 1m und 0,01 m Wasserstoffaurichlorid berechnet, findet man 0,74 lo, bzw. 45 °/o. Kout- RAUSCH’) gibt an, dass die molare Leitfähigkeit bei 18° für 1m H Au Cl, gleich 410 und für 0,01 m HAuCl, gleich 560 ist. Da die molare Leitfähigkeit von Salzsäure bei 18° in 1 molarer Lösung 310 und in 0,01 molarer Lösung 370 ist, so würde man unter der Annahme, dass das unhydrolysierte Wasserstoffaurichlorid ungefähr wie Salzsäure leitet, in 1 molarer Lösung eine grössere Hydrolyse erwartet haben, während für die 0,01 molare Lösung die gefundene Zahl ganz gut stimmt. _ Auch für Lösungen von Kaliumaurichlorid kann man mittels der gefundenen Hydrolysenkonstante die Hydrolyse berechnen, Wenn der Hydrolysengrad x und die molare Konzentration der Lösung C genannt werden, erhält man zur Bestimmung von x die Gleichung: D) ty 56 m. 1— x HITTORE ?) schliesst aus seinen Überführungsversuchen mit Kaliumaurichlorid, dass die Zersetzung des Doppelsalzes in Goldchlorid und Kaliumchlorid schon in ca. 0,4 molarer Lösung messbar und in ca. 0,07 molarer Lösung gar wesentlich stärker ist. Für C— 0,4 und für C = 0,07 berechnet man in Übereinstimmung hiermit aus der obenstehenden Formel x = 0,07, bzw. x = 0,21. 1) Zeitschr. physik. Chem. 33, 257 (1900). 2) Pogg. Ann. 106, 523 (1859). ge NE EEE me ee Zusammenfassung. 1. Das Aurorhodanid ist in fester Form, wie bereits von CLEVE nachgewiesen wurde, nur in Form von Doppelverbindungen beständig. Wir haben ausser den CLEVE’schen Kalium- und Silberdirhodanoauroaten ein neues Doppelsalz von demselben Typus, das Ammoniumdirhodanoauroat, NH,AuRh,, dargestellt. Die entsprechende Wasser- stofiverbindung, die Dirhodanoauroatsäure, lässt sich aus wässriger Lösung mit Äther extrahieren und lässt sich aus der ätherischen Lösung beim Eintrocknen als teilweise zersetztes Öl gewinnen. In seinen wässrigen Lösungen ist das Aurorhodanid immer als Dirhodano- auroation, AuRh,, vorhanden. Die Komplexitätskonstante dieses Komplexions ist 102%*mal kleiner als die des Dicyanoauroations. 5 [Au Rh, ] [Au (CN), ] i Kon ee c= — = ik 7 [Au TION PF’ Km Kran = a ; ==) L024, a [Au*]-[Rh } Das Dirhodanoauroation ist in rhodanionenreicher Lösung bei pz = 9 noch nicht merkbar hydrolysiert; in stark alkalischer Lösung wird es zersetzt. Aus den Dirhodanoauroaten entsteht durch Einwirkung von Ammoniak nach einer reversiblen Reaktion das schwerlösliche, wasserempfindliche CLEVE’sche Monam- minoaurorhodanid, AuNH,Rh. AuRh, + NH, = AuNH,Rh-+ Rh . Das Dirhodanoauroation und die aurorhodanidhaltigen Verbindungen sind farblos. 2. Aurirhodanid. Wir können die CLEve’sche Angabe bestätigen, dass Aurirho- danid in fester Form nicht existenzfähig ist und haben zwei neue Verbindungen vom CLEVE’schen Tetrarhodanoauriattypus, das Natriumtetrarhodanoauriat, NaAuRh,, und die Tetrarhodanoauriatsäure, HAuRh,,2H,O, dargestellt. Die letztgenannte Säure ist aus wässriger Lösung mit Äther extrahierbar und scheidet sich aus ihrer mit Natriumsulfat entwässerten, ätherischen Lösung mit 2 Molekülen Wasser aus. Die Bestimmung von Aurirhodanid neben Aurorhodanid kann durch Titrierung mit Sulfit bis zur Farblosigkeit ausgeführt werden. In seinen Lösungen ist das Aurirhodanid hauptsächlich als Tetrarhodano- auriation, AuRh,, vorhanden. Die Komplexitätskonstante dieses Ions ist 10177 mal grösser als die des Tetrachloroauriations. D. K, D. Vidensk. Selsk. Skr., naturvidensk. og mathem, Afd., 8. Række, V. 1. > 10 74 [AuRh,] Au re A De ; Rh 10177, [Aut**]. [Rh] * [Aut*4).[crp#’ Ka Kr = Das Tetrarhodanoauriation ist stark rotbraun gefärbt. Bei À — 578 yy ist sein molarer Extinktionskoeffizient 108; bei A — 546» ist der Koeffizient 2,6mal grösser und bei 436 wy ist er fast 10mal grösser. Bei Rhodanionenkonzentrationen grösser als 0,2 sind Penta- und Hexarhodano- auriationen in bedeutender Menge vérhanden. Die Komplexitätskonstanten dieser Komplexionen sind die folgenden: [AuRh, |] [AuRh, |] Kanth — Tan Re Je [RE PAR [ane ed ON Die Penta- und Hexakomplexe sind stärker gefärbt als die Tetrakomplexe. Bei À — 578 un ist der molare Extinktionskoeffizient des Pentakomplexes ca. 218 und der des Hexakomplexes ca. 248. Bei kleinen Rhodanionenkonzentrationen und Wasserstoffionenkonzentrationen scheint das Tetrarhodanoauriation einer Hydrolyse nach der folgenden Gleichung zu unterliegen: Au Rh, +H,0 = AuRh,(OH) + H*+ Rh. Diese Hydrolyse ist doch selbst bei [H*]-[Rh ] — 0,01 noch kaum grösser als 20°/o. Sehr bedeutungsvoll fiir das Verhalten des Aurirhodanids ist die Dissoziation des Tetrarhodanoauriations in Dirhodanoauroation und freies Rhodan, Rh,: AuRh, = AuRh, +Rh,. Die Dissoziationskonstante hat bei 18° den folgenden Wert: — [AuRh,}-[RA,] _ 10-4: K = Turn] — 049108; sie steigt ungefähr 8 °/o pro Grad. 3. Das freie Rhodan, Rh,, ist als ein zwischen Brom und Jod liegendes, farbloses, zusammengesetztes Halogen aufzufassen. Es wird aus Rhodaniden von Brom sogleich freigemacht und macht selbst momentan Jod aus Jodiden frei. Es ist äusserst unbe- ständig und wird in wässriger Lösung nach der folgenden Gleichung hydrolysiert: 3Rh,+4H,0 = 5H*+5Rh—+ HCN + H,SO,. Dieser Bruttoprozess ist das Resultat folgender 4 Partialprozesse: Rh, + H,0 = H*+ Rh + HRhAO; 2HRhO— H*+ Rh + HRAO,; H RhO + H RhO, > H*+ Rh + HRhO,; H RhO,+ H,0 — HCN + H,S0,. 75 Der 2. von diesen Partialprozessen bestimmt durch seinen langsamen Verlauf die Reaktionsgeschwindigkeit, fiir welche die folgende Gleichung als giiltig nachgewiesen wurde: d[Rh,] 72 dt [Rh,]? — k'[H RhOFP = RER Bei 18° hat die Geschwindigkeitskonstante k einen Zahlenwert von ca. 5, und ihr Wert steigt ca. 11°/o pro Grad. 4. Die Autoreduktion. Aurirhodanid wird in Lésung durch Autoreduktion nach der folgenden Gleichung in Aurorhodanid umgewandelt: 3AuRh, +4H,0 = 3AuRh, +5H*+5Rh + HCN + H,SO,. Die Anwesenheit von Wasserstoffionen, von Rhodanionen und von Aurorhodanid wirkt auf die Reaktionsgeschwindigkeit dieser Autoreduktion stark hemmend ein. Die Autoreduktion wird durch die Unbeständigkeit des durch die Dissoziation des Aurirhodanids entstandenen freien Rhodans verursacht. Es wurde gezeigt, dass ihr zeitlicher Verlauf aus dem Werte der Dissoziationskonstante des Aurirhodanids und aus der Zersetzungsgeschwindigkeit des freien Rhodans richtig vorausberechnet werden kann. 5. Die Goldausscheidung. Eine Aurorhodanidlösung scheidet Gold aus unter gleich- zeitiger Bildung von Aurirhodanid nach der folgenden Gleichung: 3AuRh, = 2Au+AuRh,+2Rh. Diese Reaktion ist reversibel, indem Aurirhodanid Gold unter Bildung von Aurorho- danid lést. Die Gleichgewichtskonstante dieser reversiblen Reaktion hat den folgenden Wert: 8 __ [AuRh,]- [Rh]? [Au Rh, ]* K = ca. 33. Die Geschwindigkeit der goldlösenden Wirkung des Aurirhodanids haben wir nicht näher untersucht. Für die Geschwindigkeit der reziproken Reaktion, die Gold- ausscheidung aus Aurorhodanid, haben wir gefunden, dass sie mit abnehmender Rho- danionenkonzentration schnell und mit abnehmender Wasserstoffionenkonzentration langsam grösser wird. Die Geschwindigkeit wird durch Zusatz von Goldpulver kata- lytisch beschleunigt. In 0,1 m HCl, 0,5 m NaRh, 0,1 m Au! wird bei 40° im Laufe von 3 Tagen ungefähr die Hälfte des Goldes ausgeschieden. Die Autoreduktion und die Goldausscheidung bewirken zusammen, dass Goldrho- danidlösungen nach genügender Zeit alles Gold in metallischer Form ausscheiden. 6. Das Löslichkeitsprodukt des Natriumaurirhodanids, [Na*]-[AuRh, ], wurde bei 18° in einer 2,2 ionnormalen Lösung zu 0,0005 und in einer 0,6 ionnormalen Lösung zu 0,0004 gefunden. Bei der Berechnung dieser Werte wurde mit Konzentra- tionen statt Aktivitäten gerechnet.!) Das Löslichkeitsprodukt des Kaliumaurirho- danids, [K*]-[AuRh, ], hat bei einer Ionennormalität von 1,4 den Wert 0,00006. >) 1) Vgl. Seite 46. Die Bestimmungen auf Seite 20 sind nur orientierend. 2) Auf Seite 18 wird 0,00008 angegeben; wenn man diese Zahl für die Anwesenheit von höheren Rhodanokomplexen korrigiert, erhält man die Zahl 0,00006. 10* 76 7. Das Aurichlorid. In einer salzsauren Lösung von Wasserstoffaurichlorid ist das Gold hauptsächlich als Tetrachloroauriation, AuCl,, vorhanden. Dieses Tetrachloroauriation ist in seinen Lösungen nach der folgenden Gleichung hydrolysiert : AuCl, + H,0 = AuCl,(OH) +H?*-+£Cr. Seine Hydrolysenkonstante hat den folgenden Wert: _ [AuCl,(OH) ]-[H*]- [Cl] _ = [Au] = 0,55 - 10%. 8. Die folgenden Normalpotentiale wurden bestimmt: Elektrodenreaktion i Normalpotential 0ËR Au+2Rh = AuRh, +0 0,689 Volt AuRh, + 2Rh = AuRh, +20 0,645 — Au+4Rh” = AuRh, +30 0,660 — Au+4Cl” = AuCl, +30 1,001 — 2hhs = Rh, +26 0,769 — Diese Werte gelten für ca. 18°. Bei ihrer Berechnung wurden die Ionenkonzen- trationen statt der Ionenaktivitäten angewandt. 9. Der von ABEGG und CAMPBELL angegebene Wert 1,5 fiir das Normalpotential Gold-Auroion ist unzuverlässig, da bei der Messung von Goldelektrodepotentialen in starker, mit Aurooxyd gesättigter Salpetersäure nur das Oxydationspotential der Salpetersäure gemessen wird (vgl. Seite 5). Diese Arbeit wurde in den chemischen Laboratorien der Universität und der königl. tierärtzlichen und landwirtschaftlichen Hochschule zu Kopenhagen ausgeführt. Herrn Prof. Dr. E. BïrILMANN, der uns in dem von.ihm geleiteten chemischen Laboratorium der Universität Platz und Material zur Verfügung gestellt hat, bringen -wir hierfür unseren herzlichsten Dank. Der eine von uns (Niels Bjerrum) hat weiter dem Carlsbergfond für eine ihm zuerteilte ökonomische Unterstützung während der Ausarbeitung dieser Abhandlung seinen herzlichsten Dank zu bringen. INHALT I. Ubersicht Feste goldrhodanidhaltige Verbindungen S. 3; Darstellung von goldrhodanidhaltigen Lösungen S. 4; Zusammensetzung und Komplexitat der farblosen Aurorhodanidkomplexe S.4; Zusammensetzung und Komplexität der rotbraunen Aurirhodanidkomplexe S.6; über die Unbeständigkeit der Gold- rhodanide S.7; die Autoreduktion S.8; die Dissoziation des Aurirhodanids zu Aurorhodanid und Rhodan S.9; die Deutung des Verlaufes der Autoreduktion S. 10; über das freie Rhodan S. 12; über die Berechnung von Oxydationsgeschwindigkeiten aus Oxydationspotentialen S. 13; die Goldausscheidung S.13; über die Deutung der Eigenschaften der Goldrhodanide mit Hilfe der gewonnenen Kenntnisse S. 15; Erklärung einiger Anomalien S.16; über die Anwendung der Gold- rhodanide zu Tonbadern S. 17. II. Darstellung und Eigenschaften der in fester Form gewonnenen, goldrhodanid- BEER EUR A D DE Te. «(five era ee ee ns ee 18 1. Die krystallisierten Tetrarhodanoauriate S.18. 2. Die krystallisierten Dirhodanoauroate und das Monamminoaurorhodanid S. 21. II. Zusammensetzung und Komplexität des Aurorhodanidkomplexes............... 24 Potentialmessungen nach Bodländer in Aurorhodanidlösungen S.24; das Normalpotential Gold- Aurorhodanid S.26; die Komplexität des Dirhodanoauroatkomplexes S.28; die Hydrolyse des Dirhodanoauroatkomplexes S. 29. IV. Zusammensetzung und Komplexität des Aurirhodanidkomplexes............... 30 Uber die Schwierigkeiten bei der direkten Messung von Aurirhodanid-Gold-Potentialen S. 30; die Messung des Aurirhodanid-Aurorhodanid-Potentials S. 31; die Zusammensetzung des Auri- rhodanidkomplexes S. 32; das Normalpotential Aurorhodanid-Aurirhodanid S. 33; die Berechnung des Aurirhodanid-Gold-Potentials S.34; die Komplexität des Aurirhodanids S. 34. V. Die Dissoziation des Aurirhodanids in Aurorhodanid und freies Rhodan....... 35 Messungen von Platinelektrodenpotentialen in Aurirhodanidlösungen S. 35; spektrophotometrische Untersuchungen über die Farbe von Aurirhodanidlösungen S. 37. VI. Über die Existenz von anderen Aurirhodanidkomplexen neben dem Tetrarho- ERR REESE REDE ES ESS EB RE CE Saren 42 Die Bildung von Pentarhodanoauriatkomplexen und Hexarhodanoauriatkomplexen bei hohen Rhodanionenkonzentrationen S.42; die Hydrolyse des Aurirhodanids S. 47. De AR FORET Sah 0042.00 nen nee ee LE TET øen SN 48 Die Oxydationsprodukte des Rhodans S. 48; über die Titrierung von Aurirhodanid mit Sulfit- lösung S. 50; Geschwindigkeitsmessungen S.51; Geschwindigkeitstheorie S.53; die Prüfung der Theorie S. 54; die Abhangigkeit der Geschwindigkeitskonstante von der Rhodanionen- und der Wasserstoffionenkonzentration S. 55; die Geschwindigkeit der Autoreduktion bei sehr kleinen Gehalten an Rhodanionen oder Wasserstoffionen S. 57; die Geschwindigkeit der Autoreduktion nach spektrophotometrischen Messungen S. 58. Me Die Zersetzungsgeschwindigkeit des freien Rhodans ............................. 60 SCO ee ete se van ou ee soie Aime ele) ole al eee cie EEE 63 Die chemische Gleichung des Vorgangs S. 63; das Gleichgewicht Aurorhodanid-Aurirhodanid an einer Goldoberflache S. 63; die Geschwindigkeit der Goldausscheidung S. 65; das Verhalten der Aurirhodanidlôsungen S. 68. eRe UN DIE DES ALE UC UNEVEN ee eue. mie eme ee essence ein, Be ele pale mwa Me ne ea © à Le 69 Das Normalpotential Aurichlorid-Gold S. 69; die Hydrolyse des Tetrachloroauriatkomplexes S. 71. A REL ED ELUES SDS Co dat dent ae a Ge RE Re er ss 72 THE LACTIC ACID BACTERIA BY Ss. ORLA-JENSEN WITH 51 PLATES D. KGL. DANSKE VIDENSK. SELSK. SKRIFTER, NATURV, OG MATHEMATISK AFD., 8. RÆKKE, V. 2 _e — OB KS + Ze KØBENHAVN HOVEDKOMMISSIONÆR : ANDR. FRED. HØST & SØN, KGL. HOF-BOGHANDEL BIANCO LUNOS BOGTRYKKERI 1919 INTRODUCTION I: development of dairy bacteriology is constantly clearing up more and more of the problems which confront the practical dairyman in his daily work. We know now that, in the great majority of cases, it is the true lactic acid bacteria which give dairy products their good qualities, while, the aerobic and anaerobic spore-formers (Ductaux’ Tyrothrix species) which were formerly regarded as being of great importance in the ripening process of cheese, not to mention the pseudo lactic acid bacteria (the Coli- Aerogenes family) are the worst enemies of the dairy industry. We are still, however, far from having arrived at a complete elucidation of all the questions involved. It is particularly difficult to understand how various sorts of hard cheese, apparently containing the same microflora,should each have its own characteristic taste and smell. There can hardly be any doubt that these sorts of cheese in reality contain different species of bacteria, only we are unable to distinguish them by the methods hitherto employed. The object of the present work is primarily to meet this want by describing the useful bacteria of the dairy industry, so thoroughly that it may be possible in the future to identify the strains encountered. A point of particular interest to the Danish dairy industry is the study of the bacteria which occur in Danish “dairy cheese” (Mejeriost), and these have accordingly been subjected to particularly detailed treatment. The material used as a starting point consisted of prize-winning exhibition cheeses from both raw and pasteurised milk (noted in the tab- les as R and P respectively). And in order properly to investigate the influence of pas- teurisation upon the microflora, we have in certain cases had cheeses made at the dairies from the same milk in raw and pasteurised state (in the tables, such cheeses are indicated by the same numbers, e. g.8 R and 8 P). On commencing the study of the lactic acid bacteria of milk and dairy produce, however, it will soon be realised that the work will become onesided unless it be extended to all lactic acid bacteria. For what we have to do is to ascertain whence the different spe- cies are derived, and how they can be found when wanted. It need hardly be said that the lactic acid bacteria of the dairy industry do not make up a complete whole in them- selves. They are introducedinto the milk toagreat extent through the cowdung, the bacteria of which, again, are to an essential degree derived from the fodder, and it was therefore necessary to study the lactic acid bacteria both of the animals and the plants. We have therefore isolated and investigated the lactic acid bacteria most frequently met with in the excrements of cows and calves, as also those of human beings, adults and children, RI» 82 4 and we havealsostudied the most important of the lactic acid bacteria occurring in soured beet slices, diffusion slices, potatoes and mash, as also in sour cabbage and sour dough. Some experiments made by the State Committee for Plant Culture, with souring of beets and potatoes, came in very conveniently for us in this respect. The strains of bacteria which we were able to isolate from these various sources were further supplemented by related forms isolated by colleagues at home and abroad. For a number of pathogenic strains, for instance, we are indebted to the colleagues mentioned in the tables, and we hereby beg to express our thanks to the gentlemen in question for their kindness in providing us with the cultures. Only 330 of the strains investigated will be mentioned in the work, as in the case of strains of the same origin, and which in the course of the investigation proved entirely identical, only one of them was included. On the other hand, we have included many apparently identical strains of different origin, which have been constantly subjected to parallel investigation in order to ascertain whether possible variations tended in the same direction or not. The lactic acid bacteria of the dairy industry are, however, those which are most numerously represented and most thoroughly dealt with, inter alia because we have had them for observation from the very commencement of the work, whereas the lactic acid bacteria of other origin only came in gradually, and have therefore in many cases only been thoroughly tested a few times, which is not sufficient to determine the constancy of their qualities. It is obvious that only the most constant qualities can be used as species character, and in investigations of this kind, therefore, all the qualities in each individual strain must be tested again and again throughout a long period of years. The present work has also, for this reason, taken up fully ten years of my time; partly also that of my assistants, and I take this opportunity of thanking them — Miss Berzy MEYER and my wife, Mrs. ORLA- JENSEN for the numerous chemical analyses, microscopical investigations and photographs to which they have devoted so much skill and care"). We shall commence by describing the methods employed in cultivation and investi- gation, passing on then to description of the isolated species and their systematics, in con- nection with which, a key will be furnished for determination of the lactic acid bacteria, as also a survey showing where they are found. And finally, mention will be made of the experiments we have carried out with a view to utilising the isolated species. For practical reasons, such of our microphotographs as it was found advisable to reproduce have been collected in a special album, and I beg to offer my best thanks to the Carlsberg Fund for having rendered it possible to publish the same. I also beg the Julius Skrike’s Foundation to accept my best thanks for having defrayed the expenses of the English translation. September 1918. ORLA-JENSEN. 1) I also beg to thank Mr. J. Linn, who has assisted me for a year, and Mr. THOLSTRUP-PEDERSEN, who in spite of having worked only a few months in the laboratory here has nevertheless contributed valuable assistance to the present work. The custodian of the laboratory, also, Mr. V. JACOBSEN, I have to thank for the help he afforded us, as an experienced dairyman, in the cheese-making experiments mentioned at the close of the work. Methods of Cultivation and Investigation. a. Cultivation and Isolation. Where many hundred strains of bacteria have to be kept under observation for years together, it requires no little work merely to keep their vitality in some measure unimpaired. At first, when we were unable to regulate the temperature during the summer months, this was a serious difficulty, and it was only by experience acquired in the course of years that we were able to master the situation. And even now, certain pathogenic streptococci, as well as some divergent rod forms, still cause trouble. The method of preservation is not the same for all species, but must be adapted to their particular demands. As regards the lactic acid bacteria of the dairy industry, it might seem natural to propagate them by transference from milk to milk. Unless very low temperatures can be maintained, however, it will be necessary to repeat this process each week, as otherwise, the bacteria will be killed off by the acid formed. We have therefore only employed this method for the more specific milk bacteria. If chalk be added to the milk, and the cultures shaken regularly, the bacteria can retain their qualities unaltered for several months. This, however, necessitates the use of large flasks, as the chalk cannot be shaken up in ordinary test tubes, so that method is only practicable where but a few cultures have to be dealt with. The employment of large flasks, moreover, increases the danger of infection from the air during inoculation. We have used this method to preserve the power possessed by certain species of forming slime in milk; a quality which is very soon lost on solid sub- strates. For preservation of the bacteria we have as a rule employed agar stab cultures in FREUDENREICH flasks. Such cultures present, in reality, the same advantages as milk mixed with chalk, as the acid, which is only formed in the stab, is thence diffused into the remaining agar mass. Although the true lactic acid bacteria only exceptionally grow in sugar-free broth, they will nevertheless all thrive in sugar-free agar!), which shows that the agar, after sterilisation, contains some soluble carbohydrates. The growth is furthered, however, by the addition of sugar, though in the case of keeping cultures, this should be restricted to 14 % grape sugar, thereby prevent- ing the formation of more than Y,% acid. In order further to keep down the hydrogenion concentration, the substrate should be rich in buffers such as amino-acids and phosphates. Prior to sterilisation, the substrates were neutralised as exactly as possible to litmus paper. For nitrogen food, we used at first 2°, 5) The growth of the sugar-loving thermobacteria is. however, extremely slight. 84 6 WITTE peptone (W) answering to abt, 0.3% N. As will subsequently be seen, this cannot compare with casein peptone (C) i. e. peptonised casein). In certain cases, yeast extract (Y) i. e. autolysed press yeast?) proved the best source of nitrogen. It is somewhat dark in colour, however, and we therefore used, as far as possible, C-Agar — and this, it should be noted, with 0.5% N,i. e. with the same quantity of nitrogen as in milk. WITTE peptone, which evidently consists for the most part of albumoses, forms strong deposits with acid, so that the stab is no longer visible in old W-Agar cultures of powerful acid formers; the nitrogen sources I have suggested, on the other hand, are free from this disadvantage. They are, moreover, rich in phosphates. A further 2%, dibasic potassium phosphate, and 1 °/o9 magnium sulphate is, however, added. A slightly larges or smaller quantity, of common salt does not asa ruleaffect the growth of lactic acid bacteria; this point will be further referred to later on. For preservation of the bacteria cultures, we have also tried cane sugar solution and other sugar solutions at various concentrations. Such solutions, which have been employed with great success for the preservation of yeast cells, are as a rule not suited to bacteria. An exception, however, is water with 2% soluble starch, which is gradually coagulated by the action of acid. To 10 cm? starch mixture is added 1 cm? of the precipitate from a broth culture. The starch solution, however, cannot as a rule compare with C-agar, as far as concerns the preservation of true lactic acid bacteria, but is on the other hand to be preferred to bacteria with surface growth (Coli-Aerogenes bacteria, micrococci, etc.) and in particular to strong ammonia-formers, which rapidly die off in the highly nitro- genous C-agar. No less important than the composition of the nutritive substrate is the temperature at which the cultures are preserved. When kept on ice, however, the air is very liable to become so moist as tofurther the formation of mould’) and we therefore restricted ourselves to water cooling. As long as the temperature is kept below 18°, the great majority of bac- teria can be preserved — without transference — for several years on the nutritive sub- strates mentioned. To make sure, however, we transferred the strains investigated every month, or every alternate month, according as they had proved more or less difficult to keep alive. Bacteria, it should be noted, are very tricky things to deal with, and it was found more than once that a bacteria strain suddenly weakened or died, while the same strain under apparently identical conditions remained unimpaired for a long time. Any 1) 100 gr. (sugar-free) acid-casein was digested for a week at blood temperature with 1 liter water, containing 4,6 °/o HCl and 2 gr. pepsin. The solution formed contains, after neutralisation, sterilisation and filtration, abt. 1 °/o N and 1.2 °/o NaCl. 2) At 50°, the digestion is completed in 24 hours. The sugar disappears at the same time. The highly acid solution contains abt. 2 °/o N. 5) As this laboratory is also used for teaching purposes, and a great deal of work is done with mould fungi, we have at times had great difficulty in avoiding infection by mould. As the mould formation often proceeds from the labels, these were always moistened with sublimate solution, before being stuck on, and the FREUDENREICH flasks were of course sealed with sealing wax. When an agar culture exhibits mould on the surface, it may easily be cleaned by letting it stand a moment with spirit over. The surface, which is then no longer dusty, is peeled off, in a layer not too thin, and the inoculation can then be made from the lower part of the stab, which is always free from mould. a 85 sudden rise-of temperature may, it need hardly be said, prove fatal to the cultures, but we have also seen cases(e.g. with Bacillus bulgaricus) where a sudden fall in temperature, of only 5°, occasioned a serious weakening. In order to determine whether the bacteria in the keeping cultures were still in pos- session of their full vitality, we compared the rate of growth with thatinfreshly reinoculated cultures. In the case of lactic acid bacteria growing in milk, all that is needed is to deter- mine the length of time which elapses before the milk coagulates, and determine the quan- tity of acid formed when this has attained its maximum, which will always be the case after 14 days at 30°. The method, however, is not without its sources of error. Even where the sowing out is done throughout with the same needle, it is impossible to avoid sowing more cells at one time than at another, and what is worse, the cells themselves may exhibit marked individual differences. Consequently, as will be seen from Table I, we do not always find, as might be expected, an even decrease of vitality in course of time. For producing pure cultures of lactic acid bacteria and, investigating the same, we likewise used substrates containing the mentioned nitrogen sources and salts. They dif- fered from the keeping substrates only in containing more sugar. The sugar broth to be used for determination of the quantity of acid formed should thus contain at least 2% sugar as some of our strains can form over 1.5% lactic acid therein. It is also less easy to over- look any development of gas when the substrate is not too poor in sugar. A slight devel- opment of gas is best observed by close sowing in tall agar tubes (Burrr's tubes). By using certain species of sugar and litmus for the plates, we succeeded in isolating the species which could ferment the sugar species in question. In order to isolate as far as possible all the lactic acid bacteria found in a starting material, it was sown out both in tall agar tubes and on gelatin plates, so that both aero- bic and anaerobic forms might develop. Besides neutral, sugar-containing gelatin (S. G.), we also used alkaline sugar-free gelatin (A. G.) on which only certain species of lactic acid bacteria grow strongly. We also had recourse to various enrichment methods. In order to get at those which thrive at particularly high or low temperatures, the starting material was added to sterile milk, and left to stand for a time at the temperatures in question. The heat-resisting species of the milk we obtained by pasteurising it in various ways, and its acid-resisting species by adding greater or smaller quantities of acid (lactic acid or acetic acid) before placing to stand at the respective temperatures. The raw cultures obtained at the first spreading were cleansed by continued spreading until all the colonies showed the same appearance and contained uniform cells. That the pure cultures thus obtained really were pure cultures was apparent from the work itself, which consisted in a constant checking of the isolated strains. As long as these remained uniform year after year, and in particular, constantly exhibited the same negative qualities, there can be no doubt as to their purity. Single cell cultures are only practicable where but a few species of bacteria have to be dealt with; not, as in the present case, where the object is precisely to study the greatest possible number of strains. As a matter of fact, we had really no practicable method of obtaining single-cell cultures when I commenced the work in 1907; it was not until 1914 that GERDA TROILI-PETERSSON published a modi- fication of the Burrı Indian ink method, suitable for lactic acid bacteria’). 1) Centralblatt f. Bacteriologie, Il. Abt. Bd. 42, p. 526. 86 8 Table I. LAM | Imme- | After | After | After | After | After | | diately ‘> year, 1 year ['/syears|) 2 years | 3 years | | Ree B , | | CE FE RE CES BES 55 Tabs | Species of bacteria No.) Preserved as So 8 FE HE 53 NEO: | | | in 30900 33 9/00 ET “loo 35 ‘/00 ge 9160 ES 9/60 | ies Bald CE Bed ER acid Es acid CE acid ER acid ss | 8 fe) jf 58 og og SE 9 à 8 log os |© i} | Se og ioe oR " | | xıv | Streptococcus lactis ........ 8| C-agar 2 54 2 \ 79,9)! 2 52| 2) 54] 2| 61) 3 5,6 U » Be eee 14 | » | 1188| 2| 741 TA LT kage 6,6 { | » cremoris: +... 11 » 1. 6,8| 2| 5,9} 2 5,2| 4) 4,5] 6| 43 mouldy XV » ime SE 18 » 1| 7,0] 2} 63| 2 6,81) 271 75,9 1220 » | | » bat HASLE »||Starch water| 1| 7,0} 2|6,8| 21:| 5,9| 2! 61] 2) 7,0) 3 4,5 el » thermophilus. | 2 C-agar 178,1 51 G2 dead 5 » 2 . | »||Starch water| 1| 81| 2| 6,81 21/2) 5,9] 2) 61) 3| 7,0 mouldy XX | » FÆDRE ee 8 C-agar 3| 6,8| 3| 5,6| 34lo} 5,6) 3) 56] 4} 56] 3 4,7 XXIV » pyogenes..... 10! » 5| 47| 6| 4,5) 6 4,3| 6/43] 8| 3,8 mouldy XXI » glycerinaceus . | 4 » 4| 5,0) 5| 4,5) 31e] 4,5) 51 62) 7| 411 6 4,1 XXII » liquefaciens... | 1 > A gl es (ac 0 ET 6,5. 217.00 1 2,00 ale as ec | Betacoccus arabinosaceus... | 9 » 10 | 4,1 3,5 3,4 2,0 1,6 dead 7 A » » ... | »|Starch water |10| 4,1|11| 3,8 2,3 3,6 2,3 1,4 Sam {| Tetracoceus') casei..:...... 5) C-agar dll 16,1 A 7, 4,5| 1.1.39 mouldy = \ » D HIVER Le 7.| » 5| 4,1) 2,712, | 34 2,3 dead Ser (| » liquefaciens ... | 10 | » 3 | 3,8 | 2) 3,2] 3 3,8| 2127 1,8 | dead ØER \ » » RE » ||Starch water| 3| 3,8| 3| 2,713 2,9| 2127 2,0 mouldy SENT N | » Pyogenes aureus |13| C-agar 3| 3,6| 5| 3,510 3219| 2,9 dead Fe \| » » » || Starch water| 3| 3,6 10| 2,0 |12 2,5|10| 1,8 1,108 2,0 ee | > > albus. |29| C-agar 6| 4,1) 27.10, 28% 2,7 2,0 2,0 3 \ » » » . | »|Starch water | 6] 4,1/10| 3,211 3,2. 14) 3,2 | 2,3 2,0 ASE ' » mycodermatus . | 31 C-agar 10 2,7 1,8 2,0 0,9 OT dead ON \ » » . | »| Starch water |'10| 2,7 | 1,6 2,0 1,6 1,4 mouldy | Streptobacterium plantarum | 3 C-agar 13-52 tt 2,3 2,0 | ht 2,5 | > N ROZ » 9! 5010| 45 3,2 2,0'| 1028 2,5 xxx | » ....112 » 10| 60/11] 56) 7 | 63/11] 4,3 2,7 mouldy » | » 6| 9,5} 5] 86) 5 | 101] 5| 9,5!) 7] 7,9) 5 6,8 j » ....| »|Starch water| 6| 95| 8| 6,110 5,6| 8| 6,3] 8| 72/10 4,5 » sus FS *'C-agar 3112,2| 311,5 | 91, 11,5) 3/12,4| 4) 124) 5| 11,3 casei... .. 2| » 7111,7| 7/11,0] 6 | 11,0] 8| 7,2| 8) 7,7} 5] 11,0 » of, Re » | Starch water | 7 /11,7|14| 8,3 3,8 | :91129,5 | 610759 4,1 > SAR 4! C-agar 4112,2| 4111,7| 6 9,0! 51104! 5| 11,0 mouldy XNIX ” eng de 11| » 4114,0| 6/12,4| 31/9! 13,5| 4114,2| 7| 8,8 » » ve 24 | » 4112,1| 4/10,6] 31}: 12,2| 3 111,9| 4] 10,4] 4| 11,3 » Sn Kenn | 28 | » 5103 |,6/"85| 9. |) 721 7172| SET INR 106 se 34 » 2116,0| 415,8] 3) 12,4] 4 1158| 4/133) | 1,6 AXVII | Thermobacterium laelis.... | 9 » 11122| 110,8} 2 9,9} 2108| 6| 74 dead ') By Tetracoccus is meant acid-forming micrococci and sarcinæ. = b. Methods of Investigation. In order to identify a micro-organism, the following points must be determined: I. Its morphological and cultural features. II. What sources of energy and nutritive matters it can utilise. Ill. Its manner of utilising the same. IV. Its attitude toward different temperatures. V. Its agglutination and other possible specific qualities. I. Morphological and Cultural Features. Of the three principal morphological qualities of bacteria: the arrangement of the flagella, the shape of the cell, and the spore formation, the arrangement of the flagella is, as I have shown in „Hovedlinierne i det naturlige Bakteriesystem’’, of primary impor- tance). The shape of the cell, on the other hand, is merely a generic character, and we may, within one and the same family, encounter sphere, rod and screw forms, as is well known in the case of the red sulphur bacteria, and as we shall also find from an example among the lactic acid bacteria. As all these bacteria appear immotile, and do not form spores, we have, in reality, no morphological indication whatever to go upon when placing them in the bacteria system. On the other hand, their biological qualities are so pronounced, inasmuch as they require just as complicated nourishment as animals, that there can be no doubt as to their place, viz. among the order of Peritrichine. A very important distinctive feature in lactic acid bacteria, and one which separates them sharply from the coli and aerogenes bacteria is that they are Gram positive. The first reaction therefore, to which we had recourse with the isolated acid formers was always to ascertain their behaviour in respect of the Gram staining process, Casein being Gram negative, the Gram staining method is excellently adapted to milk preparations. More frequently, however, the simpler staining with methylene blue is employed. This method is likewise used for demonstrating the presence of volutin grains which are frequently met with in several of the rod-shaped lactic acid bacteria. The grains are thereby stained dark blue as a rule, but at times also red. Fuchsin is not suitable for milk preparations, as it colours the casein as strongly as the bacteria. In demon- strating the formation of capsules in milk, however, this is an advantage in itself, as the unstained capsule then appears very distincly, and we have thus succeeded in showing that all lactic acid bacteria form capsules at their first stage of develop- ment. In some few strains, this capsule can attain very considerable size, but in most, it soon disappears without preliminary swelling or slime formation. The faculty of forming slime in milk is more frequent in some species than in others, but it is highly variable, and therefore absolutely inapplicable as a species character. 1) Die Hauptlinien des natürlichen Bakteriensystems. Centralblatt f. Bakt. II. Abt. 1909, XXII, No. 11—13. In this work it is pointed out that on the other hand the mere fact of a bacteria possessing flagella or not is of no systematic importance, and we also find closely related species (as for instance hay and anthrax bacille, coli and aerogenes bacteria) of which one is motile and the other immotile. D. K. D. Vidensk. Selsk. Skr., naturvidensk. og mathem. Afd., 8- Række, V. 2 12 88 10 In staining lactic acid bacteria, their acid content cannot always be disregarded. Methylene blue, for instance, will not colour highly acid broth cultures at all. Instead of adding alkali to the staining material, which may easily overcolour the preparation, it is better to neutralise the culture employed. In making Indian ink preparations, which give the best microphotographs, such neutralisation is also necessary in order to prevent the colloid ink from flaking. The illustrations in the album are, where not otherwise stated, invariably Indian ink preparations, magnified 1000 times; 500 times by means of the microscope, and twice again by the camera. We have therefore considered it superfluous to note the size of the lactic acid bacteria in the text; as a rule, they are from 0,7 to 1 „ think. A few small rod forms are exceptions; these we have called microbacteria (genus Microbacterium, PI. XLIX—-L) from their small size. It should be noted, however, that the apparent thickness of the bacteria is affected by the thickness of the layer of Indian ink, wherefore the latter should be applied as thinly as possible. In photographing and development also, the dimen- sions may be affected, so that unless the microphotographs are constantly subjected to careful control, they may turn out entirely misleading. When dealing with coloured pre- parations, it should be borne in mind that only the protoplasm, without the cell wall, is visible. Only capsule preparations give a thoroughly correct impression of the size of the bacteria. These preparations are generally photographed in water, while the other preparations are generally made in Canada balsam. A more detailed description of the morpholigical features can only be given when dealing with the separate species. The variety of forms within the present group is in reality so great that nothing can generally be said beyond what has already been mentioned. A glance at the album will show us streptococci which divide in two directions (Sc. cremoris Nr. 20, Pl. VIII, Be. bovis No. 46, PI. XXIV) or form rods (Sc. thermophilus No. 2, PI. XII, Be. bovis No. 33, Pl. XXIII) and rods, which form streptococcus-like chains (Sbm. casei No. 9, Pl. XX XVII, No. 28, Pl. XX XVIII, Sbm. plantarum No. 44, Pl. XLV, Bbm. breve No. 3, Pl. XLVII) or more or less distinct screws (Sbm. casei No. 2, Pl. XX XV, No. 33, Pl. XX XIX, No. 34, Pl. XL. Sbm. plantarum No.1, Pl. XLI and Bom. breve No. 8, PI. XLVII). These transition forms best show how futile it would be to take the shape of the cell as a basis for dividing the lactic acid bacteria into groups assignable to altogether _different places in the bacteria system. The biological qualities will show that certain spherical and rod forms (e. g. streptococci and streptobacteria or betacocci and betabacteria) are at least as nearly related one to an- other as are the spherical forms orthe rod forms respectively among them- selves. The true lactic acid bacteria form altogether a family as natural as could be desired. But if this — like other natural bacteria families — cannot be broken to pieces because it contains species with more or less long, or more or less curved cells, as has hitherto been done in bacteriology, the old generic names, which are purely morphological terms, can likevise no longer be used without some supplementary prefix, or at any rate not without giving them a new meaning. As regards the cultural features of the lactic acid bacteria, we can likewise be brief, With the exception of the tetracocci (the acid-forming micrococci and sarcinæ) and cer- tain microbacteria, they thrive best without air, and have therefore, in stab cultures, no 11 89 marked surface growth, but grow evenly throughout the whole stab; some rod forms, indeed, grow more strongly deeper down. And in conformity with this, we find the streak cultures forming only a thin veil; the plate colonies are rarely more than 1 mm in diame- ter, and coagulation of milk cultures commences from the bottom. Only a single species, Streptococcus liquefaciens, liquefies gelatin, and only a couple of pathogenic streptococcus strains form colouring matter (red staining of the stab, but not of the surface). In contrast ‘to this, the more aerobic tetracocci are generally distinguished by the power of liquefying gelatin, and form colouring matter (but always on the surface only). Some lactic acid bacteria, which we have called the betacocci (genus Betacoccus) form slime in cane sugar solutions. II. and IH. Sources of Energy, Nutritive Matters and Manners of Utilising the Same. In dealing, as here, with heterotrophic organisms, sources of energy and nutritive matters must primarily be understood as meaning sources of carbon and nitrogen, and it will be correct to take each of these separately. Carbon sources. The lactic acid bacteria utilise, like other organisms, a part of their carbon nutriment to build up their cells, especially the cell walls and other non- nitrogenous substances, the bulk of it, however, serves as a source of energy, going to form lacticacid. The organic acids are only poor sources of energy for the lactic acid bacteria!), many carbohydrates and higher alcohols, on the other hand, are particularly suitable. All our strains of bacteria were tested for the pentoses: Xylose and Arabinose, the methylpentose: Rhamnose; the hexoses: Levulose, Dextrose, Mannose and Galactose, the disaccharides: Saccharose, Maltose and Lactose, the trisaccharide: Raffinose, the poly- saccharides: Inulin, Dextrin, soluble Starch, Glycogen and Gum arabic; the alcohols: Glycerin (C,), Erythrit (C,), Adonite (C,), Mannite, Sorbite and Dulcite (C,); the cyclopara- fin: Inosit and the glycoside: Salicin?). All the sugars were d-forms with the exception of the pentoses, which were l-forms, and the dulcite, which was i-form. The tables showing the qualities of the different species of bacteria do not include fermentation of glycogene, gum arabic, erythrite,adonite, dulcite and inosite. There are no lactic acid bacteria which ferment gum’), erythrite and adonite, and it is only extremely rarely that they ferment dulcite and inosite, the fer- mentation in all cases being only slight. In this respect, the true lactic acid bacteria differ from the pseudo ones, which generally ferment these substances as strongly as grape sugar. The emission of glycogene from the tables is due to the fact that this substance is fermented by exactly the same bacteria as starch. Animal and vegetable star- 1) The most suitable is lactic acid, of which some species may form a little acetic acid and other products. See my work “Studien über die flüchtigen Säuren im Kase etc.” 1904. XIII, 161. 2) Dextrine, starch, sorbite, and salicin, however, have not been included right from the com- mencement, consequently, the quantities of acid formed by these will not always be found in the tables. Rhamnose could not always be procured during the war, and the last strains dealt with have therefore not been tested with this sugar. 3) A single sarcina (Tetracoccus No.11) was in freshly isolated state able to ferment a small - amount of gum, but lost this power later on. 12* 90 12 ches thus require precisely the same enzymes to produce any effect, and when we find that many pathogenic streptococci are capable of fermenting starch, and this very strongly, it is doubtless because they have became accustomed, in the host orga- nism, to glycogene?). Some experts may perhaps consider it quite superfluous to test the lactic acid bac- teria with four hexoses which they are probably all able to ferment. We did not, however, restrict ourselves to ascertaining merely whether any fermentation took place at all, but determined in each particular case the exact quantity °/)) of acid formed, in order to find out what carbon sources were preferred. Thus employed, the four hexoses are often of as great importance for species determination as the carbon sources which are but rarely fermented. It is highly characteristic of several species, for instance, that they prefer lævulose to dextrose, or in the case of others, that they find it very difficult to attack man- nose. As a rule, galactose is that of the four hexoses which is least fem- mented. As with the milk, we also allowed the inoculated sugar broth tubes to stand for 14 days at the optimal temperatures of the respective bacteria before proceeding to titration. In order to calculate the quantity of acid formed, the original acid grade of the sugar broth tube was subtracted. Each tube was given exactly 10 cm.? sugar broth. Many sugars are highly coloured by sterilisation in the nutritive liquids employed, which renders titration (with phenolphtalein as indicator) extremely difficult. The ones which colour most strongly are xylose and arabinose; then come galactose, levulose, rhamnose, dextrose, mannose, maltose, lactose and dextrin, while the other (non-reducible) sugars are scarcely affected. The more strongly the sugars are browned on sterilisation, the higher will be the initial acidity of the nutritive substrate?). Even though this, owing to the buffers present, may not influence the hydrogen ion concentration, the supply of buffers is itself reduced thereby, and thus doubtless also the quantity of acid formed. As sterilisation is indispens- able,it is consequently impossible to get an absolutely just comparison between the different sugars, even when they are all dissolved in precisely the same nutritive liquid — which of course we always did. It should be added, that the composition of the nutritive liquid greatly influences the transformation of the sugar on sterilisation. As our sugar-substrates contain no surplus of hydroxy] ions, it cannot be these which destroy the sugar; the trans- formation is, on the contrary, the more marked according as the substrates are richer in buffers. The reduction power of sugars is hardly impaired at all by sterilisation in pure water, whereas it is diminished abt. 20% in yeast extract, which is the richest of our substrates in respect of buffers. It need hardly be said that we have employed only the purest sugars, from Merck and Kahlbaum, for our investigations, and we always made sure that new consignments 1) It is generally supposed that the primary division of raffinose (melitriose) into levulose and melibiose is due to invertase, and it was also found that only saccharose-fermenting lacticacid bacteria were able to ferment raffinose, but it is by no means all saccharose-fermenting lactic acid bacteria which do ferment raffinose. *) By way of example, the degree of acidity (cm.’ NaOH to 100 em. nutritive li uid) for xylose, P y 4 q grape sugar and cane sugar, with casein pepton as nitrogen source, was respectively 23, 13 and 10, and with yeast extract as nitrogen source respectively 30, 20 and 15. Table Ila. %loo Acid 91 formed from °Ioo Dextrose as TAC TT | Source No. of | nitrogen 5 sk "Må r 2. 215 Jee 4 Ww 4,1 BA a. cu | ar » | 3,6 oe ee a weed u 20 8 | » 3,8 EIER. | 9 > 3,6 PRA i Se ees » 3,6 a SR La pdt ee 3,8 ALES RTE te | 17 | » 4,1 fæcrume. 4 9, sath » 1,0 Pol Ferg 8 | » 4,1 A 18.) % 4,1 SMAGEN TU » (4,3) LU SÅEDE RES | C (4,3) TU ETS | 18 w | (43) gee | ® C (4,3) glycerinaceus. . Last e NT EE 3,8 » Un Sik » 4,3 : 7% PRE 4,1 » | 6 | » 3,6 liquefaciens... | 1 » 4,3 » + | 3 | » 4,3 thermophilus . | 2 C (4,5) » R | oe ll » 4,1 » I +5 (4,5) » 5 | » 1,3 mastitldis..... NM 72 Ww 3, PA onthe 3 » 3,4 eremoris...... 1 N » | 1,8 » DN | » 2,9 a Lt. ne 10 » [Lie Qi BB... 11 » | 25 SE PRE 18 » 3,2 ARR | 19 » 2,0 SD ASA | 20 C 2,9 hee Aa | w 0,2 DE NS.» « C 3,6 — 6,1 | 20 2,9 2,9 2,9 3,6 | 2,1 5,0 | 7,2 3,6 3,4 2,0 | 3,6 3,6 23 2,9 3,6 | 2,1 | 100 | 150 | 200 38 | 3,4 34 | 2,7 32 102,6 || 36 | 2,7 | 34 | 274 32 | 2,7 34 | 2,9 34 | 27 OU ae 34 | 27 341,27 rw ri) 62.|-4,1.|° 34 50 | 5,0 12. \6.050° 6,9 34 | 2,7 34 | 2,5 36 |, 2,9 321738 37 | 32 4,1 | 34 20 | 1,4 | 0 224014 0 26 | dot <6 09 | 0 | 0 LT LOT 20 | 1,4 | A RER 32 | 23 32 | 25 | 26 | 16 | 33 | 23 | 28.| 18 36 | 36 | 27 31.124 38 | 36 | 25 92 14 Table IIb. Table | | ISO nee | "loo Acid formed from °/oo Dextrose No oi Species of bacteria No. | (0) SÆT NET | | | | |nitrogen | 5 | 10 | 20 | 50 | 100 | 150 | 200 | | | | | | | A | | Betacoceus arabinosaceus ... 6 W 11620 2.30 |e, tee, COR OM | » » u 7 » 1,4 2,0 2,5 2,7 3,4 2,9 XXV » » Ru > 144) 2.09) (2.00) 223 2S toes | » > ig » 16 1128 102840052102 70020 » » Lo THe Cc 27 | 41 | 45 | 4,5 | 46) 38 | 38 | » bavis dk 33 W 1,6% 164! 1,60) 21,80) DD eons | > AN EN IB, 34 » 101} 71,02) 5s) 20"), Sea 5274 | > parent to 35 » 184), 448.1 20 002,0 | 260300 | » PA RTE dE 37 » 0,9) "144048" } 0:8.) ie AE » Bi ART, nit] 140 » Lisl 416 11.25 91°2,9:1 S85) Dres » SEE ZE US ee MERGES 07% 09 140 aa) Te Telracoccus 9 ............. Bee yr 27.| 29]. 27 er |e 13 XXVII » BED RER EN 2 » 45 | 54 | 54 | 54 | 52 | 45 | » ON A AE 3 » 29) 92,91.) 2944 29 Dane XXVII À » CASEL PERS 5 W 2,0 1,8 1,8 | 1,8 | 18 | 09 » See Ne CO AE 6 » 27.) 2:39) 25 | eae a) cid XX VII / » liquefaciens ..... 9 | > Ji 1,4 1,5 1,1 0,8 0 \ > NAN? LO EE 0,9 | 14 | 0,9 | 09 | 02 | o XX VII » Pyogenes aureus 13. Cc 3,0 3,2 3,0 | 2,9 | 2,7 2,7 XXVII » » albus../ 29 | » | 29 | 27 | 2,5 | 25 | 25 | 09 XXVII » mycodermatus ..| 3 | » 2,9 | 2,9 | 32 | 2,9 | 29 | 2,9 Thermobacterium lactis..... GATE 0 02 |-1,L | 44) 14 1,4 » ae REN (3,4) | 68 | 9,7 | 11.0 | 11,0 | 104 | 54 » An TAN css 34)| 68 | 86 | 77 | 59 | 354 | 09 La) Å soa wht SANTE 34) | 63 |137 | 153 | 124 | 81 | 54 ravi TARA (ol: Er 09 | o9 | 1,1 | 11 | 1,1 | 06 » ER AR AR (8,4) | (6,8) | 13,7 | 13,7 | 104 |. 54 | 07 » Sur und | 11 Wh CNE ES REC TS ET » SÅ bate » | Y¥ . | (34) | (65) | 124 1 113 | 74 | 29% 07 x (| helveticum| 12 | W | 0 0' | os OL Nor. | eo Fe] , ; » | v | @4)| 72) | 133 | 128 | 81 | 41 | 0 XXVIII » Jugurt.!:.| 13 |) >» (8,4) | (6,1) | 10,1 | 9,9 | 7,0 | 29 | 07 XXVIIIÉ | » bulgaricum | 14 C A See by 52 5,4 54 | 2,9 0 \ N » | XE 661 77 ÆRA BET en 15 93 Table fic, Source | "oo Acid formed from ‘lo Dextrose Table P h No. Species of bacteria No. i of Tu IT Ts TPE ste ic 5 | 10 20 | 50 | 100 | 150 | 200 5 > Iso TS TES TEN | Streptobacterium casei ..... 4 | W (4,5) | 5,2 5,2 5,9 5,4 5,1 | » ie te 5 | » (4,5) | 5,4 | 5,5 | 56 | 5,9 | 5,4 | » ck 6] » || 56] 83 | 50 | 50 | 43 ee | » SÆR | G4] 68] 97 | 92] 92 | 88 | 8, » ie A 11 | w | @3)| 63] 64] 66] 6,5 | 63 | = > Se > | y | @4)| 63 | 124 | 135 | 140 | 12,8 | 11,3 3 eee. 16 | w Ao tab bah. G1: 17 65 | 63° | » SENTE 32 | » (4,5) 1,2 8,3 159 | 7,0 6,2 et, ir 33. || >» QS ee te Te RT NOES ne | ; BR. 7 , Y (8,4) | (7,0) | 14,2 | 15,3 | 15,3 | 15,3 | 12,4 | » EN 4 | w | 45); 68 | 65 | 65 | 65 | 63 | » Ae » | Y | (4) | (6,5) | 133 | 142 | 13,6 | 13,5 | 11,9 » plantarum een à 4,5 | 4,5 6,2 5,6 5,4 5,0 | | > » 3 | » | O | 265 | 54) 46 | 45 | 36 » > » YA 1264) |) G5) 1397 1402 l'101 | 86 | 038 | : > AT fw (4,5)| 66/72 | 721) 611 41 XXX: | » » | 10 | » | (4,5) | 6,1 6,4 5,6 | 3,6 2,5 3 » |» % foot 251299 101. | AOL A 10k | 101 > > [12 | Ww | @5)| 52) 65) 52] 50 | 27 | £ » | 20 | » | @5)| 50| 56 | 59 | 5,9 | 54 | 3 : » | v | 84 | G8) | 106 | 11,7 | 101 | 77 | 68 Belabacterium breve........ 4 | Ww | 1,4 18 | 45 | 5,0 | 45 | 4,1 » PE Pe ON LS Ze KÆBE PO | 6h) ba! un ; Eee 9 Sl ws OF TOS fae Ara | 32 | 3 a a= » | cadres Dee Teh TO 6s | 63 Pro XXXI » eta a Ga We | Oe ma 25 1 Zr) a7 | 25 » NES 12 LM 0,1 | 14 | 36 | 36 | 34 | 34 » EIER | 10 » | 0,6 | 1,3 | 29 | 29 | 28 | 4,1 » sr. |» € 2,9 | 45 | 8,8 | 88 | 104 | 74 | 65 XXXI : longum...,..| 32 | » | 16] 32 | 45 | 36 | 29 | 29 | 25 | Microbacterium lacticum. ... a | » 2,0 | 2,0 1,6 14 07 | 0 | 0 U » » +2 5 | » 2,7 2,0 | 1,8 1,6 0,7 | 0: «20 XXXII | » flavum. 8 | w 38 | 38 | 29 | 1,8 | 18 | 0,9 | XXXII Bacterium bifidum . ...... 2 | X (4,3) | (6,3) | 12,4 | 119) 9,7 |. 14 | 1,1 94 16 exactly answered to the previous ones, by testing them both with bacteria which we knew fermented the substance in question and also with bacteria which we knew were unable to do so. Many of these sugars, however, are very expensive, and it was therefore ordered that as little as possible of them should be used. Owing to the strong fermentation of many of our strains, it is, as already mentioned, necessary to use at least 2% of the su- gars. The question is, however, whether even this amount is sufficient in all cases — for when desiring to ascertain with certainty what sugars can be fermented, it is necessary that the test should be made under optimal conditions in every respect; also as regards the concentration of the sugar. Table II shows the acid production of several of our strains with various quantities of grape sugar. As we generally reckon the quantity of acid formed in °/o9, the quantity of sugar employed is here also expressed in °/o9. In order to see at all how the lower sugar concentrations behave, we have as a rule worked with an inferior source of nitrogen (W with 0.3% N); only where the bacteria did not thrive sufficiently on this we have used the better sources (C and Y with 0.5% N) but by this means, all the sugar is generally fermented in the lower concentrations!), so that the results here will be mis- leading. It will be seen from Table Ila, that the Streptococci are very little affected by the concentration of sugar, not until 15% (150/99) is reached — more rarely 10% — does the effect become pronounced. Most frequently, the optimum is found to lie at %, 1, or 2% sugar; only in the case of Streplococcus cremoris is it first reached at 5 or 10%. Table IIb, shows that the betacocci exhibit an even greater sugar concentration than the last-men- tioned species, while the tetracocci (micrococci and sarcinæ) resemble the majority of strepto- cocci in this respect. The sugar optimum of the thermobacteria lies at 2, 5 or 10% sugar, and according to Table IIc, most of the other rod-shaped lactic acid bacteria are much the same. An exception is formed by the microbacteria, which prefer the lowest concen- trations”). The main result of the investigation, however, is that we can without hesitation employ 2% sugar, whether we wish to study one or another species of lactic acid bacteria, as in no case will the quantity of acid formed therewith differ essentially from the quantity formed at optimal sugar concentration. The investigations mentioned apply only to grape sugar. In the case of the polysac- charides, the optimum of sugar concentration lies as a rule somewhat higher, though not so much as to be of any practical importance. The more or less complete exclusion of air may also affect the conditions here in question. Table IId. shows, by way of example, how Thermobacterium cereale (Bacillus Delbrücki) appeared under different experimental conditions. The nitrogen content in all tubes was 0.5%, and the quantity of acid formed is as usual expressed in P/oo- Even more important than knowledge of the sources of energy themselves is the knowledge of the manner in which they are utilised; it is this factor which determines 1) Where this is the case, the quantity of acid formed is placed in parenthesis in the tables. ” We have also investigated coli and aerogenes bacteria in their relation to the quantity of sugar, and found that they throve almost equally well with small or larger quantities. The aerogenes bacterla, however, form hardly any acid with !/» or 1°/o sugar, as they are able to convert this slight amount of sugar almost entirely into gas. Table Ild. No. of bacteria in | | = Table XXVIII 3 | 2 | 5 DER... .. … | 10 | 20 | 50 |100 | 200) 10 | 20 | 50 | 100 | 200! 10 | 20 | 50 | 100 | 200 — —=—=——- = — —— — = — ————— Y + maltose (as malt extract)') (9,5) 11,7 12,2 13,5 11,5 (10,4) 12,6|13,3 16,2 |14,0 | (9,5)112,2 12,6 118,5 | 12,6 M Mextroses: .:.1......... 6,1 7,9 92) 63 3,8| 6,5 10,8110,8 12,0 | 7,4) 7,0 13,3 14,9,13,1| 8,3 Y — dextrose covered with | EDT LT | 8,7} 9,5 110,4 | 6,3| 5,0] 6,0 113,3114,0 113,1 |11,0 | 5,4 113,1 (12,2 |11,5 | 11,5 whether we can reckon the strains investigated as belonging to the true lactic ‘acid bacteria or not. The fermentation products formed by these bacteria consist as a rule, besides dextro-,levo-,or inactive lactic acid only of a little succinic acid and volatile acids (acetic acid with traces of propionic and carbonic acid). Some few species can also form — chiefly from levulose a small quantity of mannite and hydrogen. As I have already, in several previous works”), described in detail the methods employed for demonstration of the mentioned fermentation products, I shall not go into this again at length. It should merely be noted, that as the acid formed is often a combination of active and inactive acids, the zinc lactates (whose water content and rotary power are used to determine the modification of the lactic acid) should always be allowed to crystallise out in several fractions. The more heavily soluble inactive salts will then cry- stallise out before the more soluble active ones, and it will thus be possible to form an estimate of their respective values. It should also be mentioned, that mannite may very easily crystallise out from an alcoholic extract of the dessicated culture. If the lævulose has not been entirely fermented, the remains of this should be removed before drying, by boiling with lime milk. A decoloration with bone black may then be necessary. JAN Smit*), who has worked more particularly with mannite, points out that, where a quanti- tative determination is required after the lime treatment, oxalic acid should be added, and, for removing any possible surplus thereof, again some chalk, as otherwise part of the mannite will unite with the lime. If there should be too little mannite present, to form the crystals easily recognisable under the microscope, then it will be necessary to content oneself with demonstrating it by means of copper sulphate and caustic soda, with which, like other polyvalent alcohols, it gives a dark blue solution). For demonstration of the separate fermentation products, it is of course always an advantage to have them in not too weak concentrations; accordingly, the cultures are given at least 4% sugar and a quantity of chalk sufficient to neutralise all acid formed, 1) In the maltose, however, the grape sugar of the malt extract is included. As the malt extract also contains dextrine, which is to some extent fermented by the bacteria here investigated, it is quite possible to get a greater quantity of acid formed than corresponds to the maltose itself. 2) I may here refer especially to my “Studien über die flüchtigen Säuren im Käse etc.” Centralblatt f. Bakt. II. Abt. 1904. XIII, 434. 3) Zeitschr. f. anal. Chemie 1914. 53, p. 473. 4) To 50 cm? liquid is added 25 cm.’ 4 x nNa0H +25 cm.’ copper sulphate (as with FEHLING's liquid). D. K. D Videnskab. Selsk. Skr., naturvidensk. og mathem. Afd., 8. Række, V. 2. 13 96 18 thus permitting fermentation of the greatest quantity of sugar. The lactic acid can be quantitatively determined when it has been liberated with sulphuric acid and afterwards extracted with ether, but if it is only desired to accertain what quantity of the sugar fermented is turned into acid, it is better not to add chalk to the cultures, as the quantity of acid can then be simply determined by titration. In the case of the true lactic acid bacteria, the quantity of volatile acids is as a rule small, and that of succinic acid generally even less; we can therefore reckon all the acid formed as lactic acid, without risk of any essential error. The calculation is thus very simple, when dealing with a hexose. It is somewhat more complicated when, as in the case of milk, we have to start with a disaccha- rid lactose, which may not only have become more or less hydrolysed by the bacteria, but where also the two hydrolysis products, the grape sugar and the galactose, may have been fermented in unlike degree. If the milk sugar be hydrolysed completely (by heating to 115° for half an hour with 4% H,S0,,) then its power of reduction is increased some 40—43°%1), and we can therefore, if such hydrolysis has been effected by the bacteria (see Belacoccus bovis No. 34, Table III) even find an increase of sugar during fermentation. It is consequently necessary to determine the degree of hydrolysis of the remaining sugar, and correct accordingly. All the experiments noted in Table III were made with the same milk, with 5.38% lactose, only in the last five experiments another milk was used, with 5.20% lactose (C,.H,0,,, 7.0) As will be seen, the remaining milk sugar is as a rule only slightly hydrolysed. In the present experiments, only a few thermobacteria (apart from the betacoccus already mentioned) were able to produce any considerable hydrolysis. As these formed over 14% lactic acid, it follows that lactase can act even in liquids with a considerable degree of acidity. In other experiments, where milk to which chalk had been added was used, it was found that not only had the majority of the betacocci (also those of the species Be. arabinosaceus) strongly hydrolysed the milk sugar, but generally also the thermobac- teria, betabacteria and strains of the species Streplococcus thermophilus, and Streplobaclert- um plantarum. The power of hydrolysing milk sugar can hardly be employed as species character, since it is met with, albeit not very frequently, in all species of lactic acid bac- teria. The hydrolysis is only strong in old cultures, especially those in which the greater part of the bacteria cells are dead, (in the culture of Belacoccus bovis No. 34, for instance, all the cells were dead) which seems to suggest that the active lactase is in reality an endoenzyme which is only given off when the cells are weakened, and that consequently, hydrolysis of milk sugar outside the cell is not a normal process at all. Lactic acid bacteria behave in exactly the same way towards saccharose and maltose, so that in all probability, the disaccharides are taken in as such, and there is thus nothing to prevent their being even better sources of carbon than the monosaccharides of which they are composed. As the differences shown in Table III between the quantity of sugar fermented and the amount of acid formed lie as a rule within the limits of possible error (especially considering that the quantity of sugar fermented cannot be corrected to entire accuracy) we must take it that true lactic acid bacteria transform nearly all the sugar to lactic acid. An exception, however, is Belacoccus bovis, and we shall shortly see that ') In milk cultures, we reckon with roughly speaking 40°/o, as the more strongly reducing glucose is often fermented in greater quantity than the less reducing galactose. ‘ Table III. Sugar Sugar | Lactic Increasing | Table | of the | | Fe Species of bacteria | No. | reduction | not | Sugar fermented acid d | | power after fermented fermented corrected formed hydrolysis | | | | "loo 9/00 | "loo lo | "loo XIV Streptococcus laclis......... 2 | 34 51,3 ZEN DJ 5,6 5,6 XV » cremoris...... 18 ee 48,7 5,1 6,6 5,6 XXII » liquefaciens... | 3 | 37 | 48,8 5,0 6,5 6,3 XXV Betacoccus arabinosaceus... 4 | 36 49,9 3,9 5,9 4,1 N et... | 33 | 39 433 | 105 10,9 56 XXVI | » el. gr. 34 B20 594 | 5,6 18,2 5,6 » D" RASE | 36 | 33 47,1 6,7 10,0 3,8 XXVII | Tetracoccus liquefaciens .... | 11 | 38 47,6 6,2 7,2 5,0 XXVII » RE 9 oh EV BE AE Fy ‘yes Wo Gell 9,0 5,2 XXVIII { Thermobacterium lactis..... | 6 | 2 In 444 oh, 34 15,6 12,6 » tt RE ih GS 442 | 96 17,1 12,8 XXVIII roy bulgaricum | 14 | 20 45,2 | 86 17,6 12,8 XXVIII » helveticum. | 12 | 40 | 34,5 19,3 19,3 19,3 Streptobacterium casei...... 2 | 35 41,3 12,5 14,6 13,1 » CE ets ses AU 038 41,6 12,2 13,0 12,8 » eee | 5 37 42,8 11,0 12,3 11,5 » ROVER GE JES 37 42,9 10,9 12,2 10,1 XXI ’ ’ 2 tj ig » N | 13 | 29 42,8 11,0 15,7 13,7 » NA = à 18 35 | 41,6 12,2 14,3 13,3 3 ee sg | 37 | 363 17,5 18,6 16,9 » its 34:11 87 37,8 16,0 17,1 16,7 » plantarum, 1 | 34 45,0 8,8 11,5 10,4 » » Ira 37 45,7 8,1 9,4 7,7 » » 8 | 37 48,2 5,6 7,0 6,5 | 1 ? 2 ? XXX > » 13 | 36 43,2 10,6 12,3 10,1 » » 15 37 45,2 8,6 10,0 Ge ? » IN AE: | 38 42,6 112 | 12,1 11,9 XXXI Betabacterium breve... KEEL 3 | 39 i 50,6 3,2 L 3.1 2,0 | Bacterium coli A1)......... 1 40 47,2 4,8 4,8 5,5 » DT) SED RE 2 41 | 48,8 3,1 3,7 4,2 » prodigiosum ..... 1 | 41 42,5 9,5 9,5 3,6 » aerogenes........ 1 19 23 | 49,7 500 | 4 » EP RARE 2 0 52,0 52,0 2,8 Belacoccus arabinosaceus behaves in a similar way with other sugars than lactose. That we nevertheless reckon the betacocci as among the true lactic acid bacteria is due to the fact that they are allied to these in all other respects. The betabacteria also, can, in a freshly isolated state, transform a great quantity of the sugar to something other than lactic 1) A and B indicates the relation of the Coli-forms to sugars according to C. O. Jensen. Both of them essentially form succinic acid, A besides some dextro-lactic acid, B some levo-lactic acid. The quantity of volatile acids (1 part of propionic acid to 10 parts of acetic acid) makes up 30 per cent of the fermented sugar. 13* 98 20 acid, but as a rule, they lose this tendency after continued cultivation in the laboratory, and will thus, in course of time, be found to differ in no essential degree from other lac- tic acid bacteria. In the system which we shall later set up for the lactic acid bacteria, we have nevertheless separated off the betabacteria and betacocei as a distinct sub-group. At the close of Table III, we have shown, for purposes of comparison, the acid pro- duction of a few other bacteria. It is highly surprising to find that the two coli strains — despite abundant air formation — apparently form over 100% acid from the fermented sugar. The explanation, however, is that these bacteria only form a small quantity of lac- tic acid, but ferment the greater part of the sugar to succinic acid and acetic acid, thereby giving three equivalents of acid for every two obtained by lactic acid fermentation. C,H,,0, = 2C,H,OCOOH = C,H,(COOR), + CH,COOH + H, 1 grape sugar — 2 lactic acid = 1 succinic acid + 1 acetic acid + 1 hydrogen. By calculating the acid formed as lactic acid, then, we have thus put the yield of acid 1/, too high. Baclerium prodigiosum forms acids similar to those formed by the coli bacteria, but as it also transforms a quantity of the sugar completly into gas, there isa distinct wastage here. Transformation into gas is, however, most marked in the caseof the aerogenes bacteria, and especially No. 2, which has fermented all the sugar in the milk without coagulating it. A point of no slight interest is the question whether a given lactic acid bacterium will under all circumstances form the same amount of by-products, especially volatile acids, and whether it always forms the same modification of lactic acid. Here, above all, the quantity of sugar plays a part, for if there is not any more sugar than the occurring bacteria can casily ferment, part of the lactic acid formed will be further transformed. This is plainly seen from Table II where the quantity of acid formed is considerably less than the fermented quantity of sugar in the cases where all the sugar has been fermented, i. e. where the figures are put in parenthesis. As regards the formation of acetic acid, KAysER!) has already shown that this increases witb the supply of air, and BARTHEL?) has pointed out that it is as a rule greater where the conditions of life (f. in. the temperature) are unfavourable. In accordance with this we have found a relatively far greater quantity of acetic acid in cultures without chalk than in those with. True, the quantity of acetic acid formed in milk cultures with _and without chalk is about the same, but as a far greater quantity of sugar is fermented when chalk is added, it follows that from the sugar fermented more acetic acid is formed when chalk is omitted than when it is added. Some examples of this are shown in Table IV. We have reckoned one molecule hexose as giving three molecules of acetic acid. The quantity of volatile acid formed by the lactic acid bacteria depends not only upon the conditions of life, but varies also in other ways. BARTHEL found, for instance, that Sc. fæcium No. 19, in a freshly isolated state, formed from the sugar fermented 39% volatile acid, whereas we, nine months later, under the same experimental condi- tions, found only 13% volatile acid. The bacterium in question exhibited no sign of weakening, but formed, indeed, more total acid in milk than it did in a freshly isolated state. Sugars whose number of carbon atoms is divisible by six do not as a rule affect the ') Contribution a l'étude de la fermentation lactique. Paris 1894. °) Centralblatt f. Bakteriologie. II. Abt, 1900. Bd. IV, p. 420, 21 | 99 relation between lactic acid and by-products. In the case of the betacocci, however, it is otherwise, as they form from lævulose, and therefore also from saccharose, more acetic Table IV. Addition Sugar Acetic Acetic acid ker Sorina ok Wecteriä No. År fermented, acid percentage RE" st chalk corrected formed of fermented 6/00 9/00 sugar i | + = = = I 7 7 Sr ar 7 RUE yo} De Lu 4 | poet 31 1,0 3 XV | Streptococcus cremoris............ | 20 | ae 7 1,0 14 XXVII | Thermobacterium lacti he oe OR x 2 ermobacterium lactis........... PR or 13 0,4 3 : FE N 52 0,8 1 XXVIII | » helvelicum...... 19, A 0 19 12 6 XXIX | Streptobacterium casei | 17 ET 2 | ne å 2 ptovacterium casel............ (|| \ 0 12 1,0 8 i ae TR; - 52 0,5 1 XXIX 5 KB Tue year Oe a ee | 18 L 0 10 0,5 5 | | + 52 0 1 | 34 | LUE 4e XXIX PRES PRES | 2 | er ED ae : acid and gas than from other hexoses. Table V shows the fermentation of levulose and grape sugar ina strain of strong gas-developing character (Belacoccus arabinosaceus Nr. 12). Table V. Sugar Lactie SEER Species of sugar | fermented acid Dex: Rest | formed formed 0/00 0/00 0/00 "loo Levulose...... | 1957 3.0 2,5 13,6 Dextrose 2/1. | 12,8 5,6 0,5 6,7 As will be seen, this coccus forms almost as much acetic acid as lactic acid from lævu- lose, but on the other hand, only 1/,, as much acetic acid as lactic acid from grape sugar. The fermented sugar which is not transformed into acid(the remainder in the last column of the table) has not all been formed into gas (carbonic acid and hydrogen) at any rate not in the case of the levulose culture, which gave a very distinct mannite reaction. Now one molecule of hexose can by the lactic acid bacteria be simply broken up into two molecules of lactic acid, whereas sugars having a smaller number of carbon atoms, will of course be otherwise divided. It would seem reasonable beforehand to suppose that pentoses would be broken up into equivalent quantities of lactic acid and acetic acid, and as a matter of fact, there is really always a large quantity of acetic acid formed when pentoses are fermented by lactic acid bacteria. As a rule, however, we obtain considerably less acetic acid and more lactic acid than would answer to the equation C; Ay, 0; > C; H, 0; + C, H,0,. 100 | 22 It will be seen from Table VI, that we have found in the case of the betabacteria a division exactly answering to the equation 6C,H,0, = 8C,H, 0, + 3 C,H, 0, 6 arabinose — 8 lactic acid + 3 acetic acid. The bacteria thus appear as true lactic acid bacteria, inasmuch as they form the greatest possible quantity of lactic acid from the source of energy given'). This is, how- ever, on the supposition that chalk is added to the cultures; without chalk, we obtain here, as in other cases, proportionately more acetic acid. Table Av Arabinose| Lactic Acetic EU hs. Ratio = aci ac en : fer- acid acid equiva Betabacterium breve No. 10 > 3 mented formede formed formed formed lent: ; em © cm3 R || Lactic acid 0100 0/00 0/00 4 4 | Acetic acid Arabinose casein pepton broth with chalk. | 35,2 28,0 7,2 125 48 | 8:3 » » » » without » . 17,8 13,1 4,7 58 31 (ZEN Up to now, we have reckoned the entire quantity of volatile acids as acetic acid and this we may safely do, as there is at the outside 1/,,, more often only 1/, propionic acid, and a trace of formic acid mixed therewith. In the case of certain betabacteria, we have already mentioned that more gas is formed with cultures in a freshly isolated state, and as regards the betacocci, the development of gas is undoubtedly in proportion to their well-being. Much would seem to suggest that the transformation of sugar in all cases takes place by way of lactic acid, and that it can only be further divided — i.e. made to yield more energy — by strains of particularly marked vitality. The majority of the true lactic acid bacteria, which do not develop any measurable quantity of gas, can, however, -— likewise when in a state of particular vitality — produce in milk so much carbonic acid that fine stripes appear in the curd. The lactic acid bacteria, then, as mentioned, form most acetic acid under unfavourable conditions, whereas exactly the reverse is the case with the carbonic acid, of which most is formed under favourable conditions. Experiences from our previous works warrant the supposition that the rotary power of the lactic acid formed constitutes an important specific character for the true lactic acid bac- teria, and we have therefore determined, in the case of all strains examined, what sort of lactic acid they formed, not only in milk, but also in broth with different carbon and nitrogen sources. These investigations we have repeated from year to year. The new investigations have on the whole confirmed the correctness of our supposition. As a rule, neither the carbon sources nor the nitrogen sources affect the modification of the lactic acid. Those strains which in milk form pure dextro- or levo-lactic acid will also in a nu- ') This is the more surprising, as the ancestors of the strains used (Bacillus y. in my thesis above quoted), formed much succinic acid, with abundant gas development. 23 101 tritive broth always form dextro- or levo-lactic acid, whether the source of energy be alcohols, aldoses, ketoses, pentoses, hexoses or polysac- charides!). Those strains which in milk form purely inactive lactic acid — i. e. with like quantities of dextro- and levo-acid — will as a rule also under other conditions maintain the equilibrium between the two acids, whereas strains which in milk form more of the one than of the other will under less favourable conditions generally only form the acid which they most easily produce. Indeed, we do not even need here to alter the nutri- tive substrate, as even in milk, these bacteria can in the course of years end by being only capable of producing the one acid. We have numerous examples, for instance, of cases where strains of the species S/replobaclerium casei, which in a freshly isolated state forms, besides dextro-lactic acid, also smaller or larger quantities of levo-lactic acid, have after a more or less considerable lapse of time been found capable only of forming dextro-lac- tic acid, and that without any decrease in the total production of acid. These investigations, then, distinctly show that the modification of lactic acid is altogether independent of the stereochemical structure of the sugars, and depends entirely upon the species of bacteria. We must therefore presume that dextro- and levo-lactic acid are formed each by its own independent enzyme. Strains which are equally supplied with both (as for instance Thermobacterium helveticum) will under all conditions form purely inactive lactic acid, whereas strains which can more easily produce one of the enzymes than the other, may often entirely lose the power of producing the latter, and thus the faculty of forming the corresponding lactic acid. Nitrogen Sources. In contrast to the pseudo lactic acid bacteria, the true ones do not thrive with ammonia salts or single amino-acids as source of nitrogen. We have in this latter respect tested al] our strains of bacteria with aspartic acid, but none of them showed any signs of growth. The true lactic acid bacteria demand just as compli- cated nitrogenous food as the animals, viz. genuine proteins or the entire complex of amino acids therein contained. Even incomplete proteins, such as gelatine (without addi- tion of other nitrogenous nourishment) generally proves, as in the case of animals, an extremely bad nitrogenous food. As the lactic acid bacteria — save for the few species capable of liquefying gelatin — do not in a living state give off any proteolytic enzymes, the nitrogenous food must be given in a state of solution or in colloid form. Of all genuine proteins, the most suitable seems to be casein in the form in which it is found in milk. Even better, however, in many cases, is paracasein (or rather, perhaps, the peptones which rennet gradually forms from casein); this will be seen from Table VII, where we have noted the quantities of acid formed by various species of our lactic acid bacteria in the same milk with and without addition of rennet. The rennet employed was rendered germ-free by filtration through a sterilised CHAMBERLAND-filter?). The effect of the rennet 1) HERZOG and HôRTH have, in their work “Zur Stereochemie der Milchsäuregärung” (Zeitschr. f. physiol. Chemie 1909. Bd. 60, p. 131) arrived at the same result. 2) As such filtration, which keeps back some nitrogenous matter, weakens the rennet very con- siderably, it is necessary to start with a very concentrated solution. We dissolved 10 Hansen tablets in 200 cm? of water, and used 2 drops of the filtrate per 10 cm.” milk. The nitrogen content of the 102 94 Table VII. | | | 900 Acid | Joo Acid | te formed | formed Ae | Species of bacteria No. ss ae | FR Species of bacteria No. on pers | |e ae] Så |s2 | (SE FEB] 28 ÈS | (sa | | ee A i | Streptococeus laclis......... 6 | 3,4 | 4,5 Streptobacterium casei...... | 11,6,1176,3 XIV | » FD, bees ARE 71143|6,31 » SL Das Uae Gi | 2) 9,9] 11,3 > Eee 141 7,4] 8,1 > Ale 411,3 | 12,8 » cremoris ..... 1} 4,7) 6,5 » Sih. Sy A 910,4 | 10,6 XV | 5 Du ee 2| 4,6 | 6,1 » peser 10 11,9 | 13,3 | » » 10| 5,6 | 6,5 » RCE - 11 114,2 | 14,4 » SPA SSRI E GE 21 6,3 152 XXIX » De 22T 14 SH 11,5 XX f » FCW SX: 14| 4,1| 5,2 » Ee oe 15113,1 | 13,1 \ » 7 2 lc 18| 4,5 | 5,6 » D? 2 oats 23 |11,0 | 11,9 xx1 J » glycerinaceus..| 4| 4,5| 5,6 » CARE 25 112,8 | 12,8 \ » » .. | 61 2,5) 4,7 » +0 ane 27 113,1 | 15,8 XXII liquefaciens ...| 1| 7,1| 7,1 » ST fe 32 114,6 | 16,0 f Betacoceus arabinosaceus...| 4| 5,6| 6,8 » OR ig nat 34 116,0 | 17,1 XXV | » » 7| 5,2| 5,4 » plantarum | 1| 5,9} 8,8 » » 8| 3,8} 5,2 » » 2|| 2,9! 5,4 xxv f » DODIS. ER et 34| 3,2| 3,8 » » | 31 32| 7,0 \ » Wilt ato Fae de 40| 0,5) 1,1 | » » TØ 3,8| 9,2 XXvq J) Tetracoccus casei .......... 5| 41| 361 XXX {| » » 91 8,3 | 10,4 \ » ER ENT te 6| 2,7| 2,7 » » 11| 3,6] 5,6 XX VII / » liquefaciens ....| 9| 2,5| 2,7 | » » 12| 5,0| 8,8 | » Dt Rios 11} 0,7 | 0,7 \ » » 14| 9,5 | 11,0 xxvint J) Thermobacterium lactis... . 6 10,4 |10,4 » » 15) 4,1| 7,4 \ » ay Aree 111,0 7e f Betabacterium breve........ 4 1,8| 1,6 XXVIII | » helveticum. | 12/19,6 |21,6 | XXXI | » AN SEAT 5| 0,9| 0,9 XXVIII » bulgaricum. | 14113,3 13,7 | » Stee care 10 LB | 2,0 | | | I is of course not perceptible if the bacteria themselves give off proteolytic enzymes, as is the case with S{replococcus liquefaciens and Telracoccus liquefaciens. It is likewise unnoticed in the case of other tetracocci, certain thermobacteria, the betabacteria, and certain strains of the species Streplobaclerium casei, but is quite extraordinarily distinct in the case of Streplobaclerium plantarum. This species, for which casein is as a rule a relatively poor source of nitrogen, forms, when rennet is added to the milk, far greater quantities of acid than otherwise. This fact throws new light upon the ripening of rennet cheese, a process which is due to the action of the rennet and of the lactic acid bacteria. We have already previously shown that the action of the rennet is furthered to a high degree by milk is thus not perceptibly affected, and the milk, being highly sterilised, does not coagulate. It does, however, coagulate, as soon as any trace of acid is formed, and after inoculation with lactic acid bac- teria therefore, milk to which rennet has been added will coagulate far more rapidly than milk without rennet. If the bacteria themselves give off proteolytic enzymes, then there will of course be no differ- ence, practically speaking, in the time required for coagulation. i 57 ; 25 | 103 the acid formed by the lactic acid bacteriat), and we have thus now shown, on the other hand, that the action of many lactic acid bacteria is furthered by the rennet. And in this connection it should be mentioned that BARTHEL has recently shown?) how consider- ably rennet increases the casein-splitting power in Slreplococcus laclic. The various factors which give rise to the ripening of cheese thus mutually accele- rate one another’s action. Of other genuine proteins which may be utilised for the cultivation of lactic acid bacteria, we may mention gluten and legumin dissolved in the smallest possible quantity of sodium phosphate. This solvent may also well be employed where casein is to be used in connection with other sugars than milk sugar. The most typical milk bacteria grow best in milk, and only with the greatest diffi- culty in peptone solutions; there are, however, lactic acid bacteria which, even though they ferment milk sugar in peptone solutions, thrive peorly in milk, or require at any rate to be accustomed to it, and which will in consequence rapidly lose the faculty of so doing if left for many generations without coming in contact with milk at all. Indeed, it seems possible to accustom the bacteria to largely dissimilar forms of nitrogenous nourishment, and it is doubtless in many cases here that the main difference lies between the parasites and the saprophytes most nearly related. The pathogenic forms have often so accu- stomed themselves to a certain particular nitrogenous food (that from which their toxins are also formed) that they can ill thrive without it. It has thus on the whole proved considerably more difficult to keep the pathogenic streptococci alive than the sapro- phytic. Many of the rod-shaped lactic acid bacteria of the digestive tract, also, were difficult to cultivate, owing to the fact that we had not succeeded in satisfying their particular requirements in respect of nitrogenous food. When seeking to compare the values of different nitrogen sources, the only possible method is that employed in agricultural chemistry, to wit, by offering the organisms the same quantity of nitrogen in the different forms, but under conditions otherwise uniform. In Table VIII will be found noted the quantities of acid (calculated, in the usual way, as %o of the nutritive substrate) formed by some of our strains in horse serum (S) Liebig’s meat extract (L), Cibil’s do (Ci), Witte peptone (W)®), casein peptone (C) and yeast extract (Y), partly with the separate sources alone, and partly when used with the addition of a 2% Witte peptone solution. It was so arranged that the quantity of nitrogen was through- out 0.4%, 2 percent grape sugar was added, and the quantity of the different nutritive salts was also kept as far as possible uniform throughout, save that no potassium phosphate was added to the Wy. The last three columns of the table show the effect of this salt upon the casein pepton broth used by us, with 0.5% N. We had expected that the pathogenic bacteria would have preferred blood serum and meat extract, and the milk bacteria casein peptone. This however, did not prove to be the case. The blood serum proved throughout a bad source of nitrogen, even for pathogenic bacteria, and this despite the fact that its bactericidal substances 1) Landwirtschaftliches Jahrbuch der Schweiz, 1904, p. 404, and 1907, p. 97. 2) Meddelande Nr. 171 fran Centralanstalten för försöksväsendet på Jordbruksomradet. 1918 8) The serum contained 1.13 °/o N, Lresic’s Meat Extract 9.22 “Jo N, and Cısır's do. 3.11°/0 N. Equal quantities of nitrogen from the two meat extracts had very nearly the same effect upon most of the bacteria used. D, K. D. Vidensk. Selsk. Skr., naturvidensk. og mathem, Afd., 8. Række, V,2. 14 104 26 Table VIIa. | | All nutritive mediums with 2 ?/, Dextrose se c. with 2° lo Dex- | and 0,4°%o N |trose and 0,5 oN eke Species of bacteria [Re | | | | | ES BE and me | | S ANA FF EEE Dee = Hin. Ti TESTEN ER mern | i | Streptococcus pyogenes...... | 1 [0,941 |1,412,3 | 2,7 18 2,312,012,3|12,3| 2,9 | 4,1 | 4,1 : TE CE | 8/o7138/14114132/02129/23|25|16| 34 | 43 | 43 XXIV a 4 10,713,6,09|09 2,5 | 0,2/1,8/1,1/1,6/1,1] 2,9 | 36 | 4,1 | » ae 7 10,9! 5,6/1,6|2,7 4,3 15,6|3,4/3,4|3,613,6) 56 | 6,3 | 63 | MATE 8 12,5 5,4 |0,5|2,7|34/3,2/2,7/3,4/3,4/29) 38 | 54| 5,6 XXIV ; "al AE SES | 11 | 1,6|5,6/0,7|3,6 | 4,7|/6,8|3.4|5,4|5,0|5,0) 5,9 | 7,9| 8,1 XX » Tec a | 8| | 7,0 | 0,7 3,8 | 5,2 | 6,8 3,61451/4,715,0) 541 7,2 | 7,1 XIV i RE x 29 4 14 | slar|65l23136152163136145147143| 59 | 79 | 7,4 [ » eremoris...... | 1 10,2 | 4,7 | 1,6 | 2,5 | 2,9 | 2,9 | 2,7 | 2,3) 2.3/2.9) 45 | 47 | 47 Xv a: | 18 }1,1| 5,2] 1,8/3,2|6,4|3,6/3,8/3,6/3,6/36! 52 | 6,5 | 7,0 | » Bl ltr ts | 19 10,2 | 3,4] 0,9 | 2,5 | 4,3 | 4,3 | 4,1 | 3,6/4,5/3,4] 43 | 5,6 | 5,9 mi mastitidis .... | 2 |25/4,5|1,6|2,7|3,4|0,2|2,9/2,5/2,7|29] 4,1 | 4,7 | 50 \ ODED | 31230,50114127|34129132|29 2923| 45 | 54 | 5,6 XVII thermophilus.. | 2|2,916,311,812,715,415,2|4,7|3,2|2,7|3,6| 6,8 | 8,1 | 86 XIX > inulinaceus ... | 4 |0,9|4,3| 1,6 |3,6| 4,5| 4,7 |4,1|4,3/3,8/ 4,1] 4,1 | 4,7 | 5,4 XXI {| x glycerinaceus.. | 4 6,3 | 2,5 |3,614,3|5,413,413,814,515,0| 5,9 | 7,7 | 8,1 \ » » … | 6129/63125/36)4,515,0/4,11/38/41141) 56 | 65 | 68 Lt 3 liquefaciens... | 1 (4,1/7,2|2,7|3,8/4,3|6,5/3,6|4,3|4,3|5,0! 5,0 | 6,3 | 7,0 \ » , ma: 6,3| 2,5 | 3,2|4,5| 5,6|2,5/3,6|4,1/3,4| 5,4 | 7,0 | 7,4 [ | Belacoceus arabinosaceus ... | 4) 1|0,7/0,9|2,9|3,4|6,8|2,9| 3,8 | 3,6 4,5 | 4,3 | 5,2 | 5,4 XXV > > 8 0,3 | 5,2| 2,9 | 2,9] 3,4 | 7,7|3,6| 4,3/4,1/ 4,3] 38 | 3,8 | 5,0 | » » | 12 | 5,2 | 2,7 |2,7|3,8|8,112,5|3,613,614.1| 4,5 | 5,4 | 5,4 xxvi | » bovis ........... | 34 /0,5| 5,4| 20 2,7] 4,5| 7,0/3,2|4,3/43/4,3) 59 | 6,1 | 6,1 aber | » x LR, sobre | 41 7,010,913213,417,212,7|14314,5145| 5,0 | 5,0 | 6,1 XXVII | Tefracoccus casei........... | 5 |, 13,8] 1,4 | 2,0] 2,7) 0,9] 1,8] 2,7/3,2) 2,7] 2,9 | 34 | 34 XXVII EN | 8 20|235|0,9|0,9]1,8|09|1,410,9 1,4/0,7/ 16 | 2,3 | 38 XXVII liquefaciens..... | 91 2,5 | 0,5 | 0,9} 1,8 | 0,5 | 1,1] 1,1}1,6/}0,9} 1,8 | 2,7 | 2,9 XXVII pyogenes aureus | 13 12,014,5120123|25|3,6)23|2,3|2,9)2,0| 3,4 | 4,3 | 4,5 XXVII > > albus.. | 99 | 1:1|3,2/0,7|0,9|20/1,8/1,4| 14/20/32) 18 | 29 | 32 had been inactivated by long standing: meat extract, on the other hand, was on the whole at least as good a source of nitrogen as casein peptone. Only Microbaclerium lacticum decidedly prefers the latter. That we have not, in the present work, employed meat extract is due, besides its dark colour, to the variability of its composition. The meat extract now generally sold is no longer mainly composed of creatin and purin bases; far from it; the overwhelming majority of the nitrogenous matter it contains is furnished by albumoses and peptones. Yeast extract exhibits a more specific action, proving an extremely bad source of nitrogen for a number of pathogenic bacteria’), (Sc. pyogenes Nos.3 and 4, and Sc. maslilidis No. 2), some few micrococei (tetracocci) and Mbm. lacticum, whereas it furthers to a surprising degree the develop- ') The sell known therapeutic properties of yeast are doubtless connected with this feature. 27 105 Table VIII b. | All nutritive mediums with 2°/o Dextrose | C. with 2 °/o Dex- Table i . _ and 0,4°/o N | trose and 0,5 °/o N No. Species of bacteria No. a” Tre Sen ARE ay A OE (| Ele - sız|w|iw|c EN er FP ae ol EUR LUN DT” ee Ye = | pete ea ac ee vies z Aa | Thermobacterium laclis..... 6| 0 4,3 | 0,5 | 2,3 | 0,7 | 10,6 | 6,8 | 2.9| 3,6 | 7,4 | 6,1 | 7,0 | 7,2 .: » AL | 8| 0 | 7,011,615,0| 0,9,15,3[9,9|4,1|8,1110,1111,5 | 86} 95 | » Brits 3, 10| 0 | 81/0,7/2,5] 0,9] 12,8 | 6,5|3,2/4,1| 4,7| 9,9 | 8,3] 9,7 2 » ..... [11] 0 | 5,9/0,715,4| 5,9] 10,6 | 5,4 5,9 | 5,6) 5,4) 86| 95] 9,0 XXVIII | » helveticum |12| 0 | 1,411,410,9| 0,7|11,313,2|3,4|3,6| G,8| 4,5 | 6,3| 7,4 XX VIII | » bulgaricum |14| 0 | 1,8/0,7|1,4| 3,4|12,2|7,0/0,7/0,9| 6,8) 2,3 1,6) 2,7 Streptobacterium casei...... 1 0 |: 7,712,713,2| 3,4115,813,813,8|4.1| 5,4| 2,3 | 3,6) 3,5 » SCALE | 2 8,6 | 4,0] 4,0| 7,0] 17,6] 2,3| 5,4/ 3,8] 7,2| 6,5 | 83] 8,8 XXIX }| » ei ts: 110,7 | 9,7| 3,6] 6,1] 8,3] 18,0 |8,1/9,0 | 8,1 111,3 10,6 | 10,8] 11,5 | | ; sm... 16/05! 8,1132]52| 68) 17,8 17,0 7,417,4| 7,7| 99 | 104) 11,5 » RE ‚343,8 (11,7 | 8,3 | 8,3 111,0] 17,6 |9,0 8,8 8,3| 8,6113,7 | 14,2| 14,4 » plantarum | 1| 0 | 86120|6,3| 6,3| 17,6 | 8,6| 6,3| 7,2 |11,0] 7,2 | 6,8| 4,5 » » 3] O | 8111,8|5,4| 7.3] 14,4|7,7|6,8}6,1| 6,8| 8,8 | 10,4) 10,6 Xxx » » 5, 81116168) 8,3|12,4 | 5,2 177181! 9,7111,0 | 12,4] 13,1 » > 111134! 9,0|3,6| 3,6) 8,3] 16,2|6,1| 6,5) 7,0) 8,0) 9,5 | 10,8) 11,3 » » 112) | 8,6)0,7|6,8| 7,2| 16,7 | 5,0 | 7,4| 7,4) 9,9) 7,3 10,1), 11,3 » > 21| 8,6 13,4|5,9| 95117,315,616,814,5| 7,4|10,1 | 11,9| 12,4 Belabacterium breve ....... 3 61134152| 5,2| 61126147|45| 5,9] 52 | 63| 6,8 XXXI » NN .. | 4118| 68123145! 36| 7014,1120|1,6| 7,0) 47 | 50| 6,1 » “où SRE 10 5,0|2,5|2,5| 52| 861217120120! 7,2] 61 | 6,5| 7,9 | Microbacterium lacticum.... | 2 24\14|14|25| 0 |1,6/114120| 2,4) 39 | 3,6) 3,8 se » » | 312,3! 0,9 1,4| 1,8) 2,9 0,2|2,5/1,8)2,3| 1,4] 32| 4,5) 4,5 | » » | 4 1,8] 1,6] 1,6| 29| 0,211,812012,3| 1,6] 2,9| 4,1| 4,5 » » | 611,8) 1,1/1,412,0| 3,2| 0,512,3123,0|2,3| 1,4 32| 43|. 43 XXXII > flavum..... | 810,5! 5,6 | 2,5|3,2| 5,6| 6,511.6,5,615,2| 6,5) 54 | 65) 7,4 Bacterium coli À ......... 1 40123123! 3,6| 3,812512,913,4| 2,9] 38) 54) 56 » paratyphi....... 6|2,7| 4,5/2,3/2,3| 36| 38123123127| 2,3] 38| 52| 5,2 » prodigiosum..... 1 a1l14 3541| 32123129129| 27| 36| 38] 3,8 » aërogenes ....... | 413,21 2,3/0,5| 2,9} 2,9] 2,511,412,9/3,2| 2,3) 0,5 | 0,9) 0 » ARRETE 512,9|1,1132|36| 5,0|+0,9| 1,1] 3,6/3,8| 2,9) 2,7 2,7|-0,5 » acidi propionici.: | 3| 0 | 9,9] 0 10,5] 3,6/13,7,0 090,7 11,5 8,1 10,4, 10,6 ment, and even more the acid formation, of the genera Thermobaclerium ?) and Streptobacterium. For the genera Belacoccus and Belabacterium, also, yeast extract is fully as good a source of nitrogen as casein peptone, whereas the reverse is the case with 1) The results in the case of thermobacteria are often somewhat uncertain, as these bacteria are so sensitive to nitrogen sources, that they do not grow they have overcome the difficulties, however, they may then suddenly grow through very powerfully. Both meat éxtract and yeast extract for instance seem to contain substances which retard the develop- ment of Tbm. bulgaricum, but it generally overcomes the retarding effect of the yeast extract after the at all if there be the slightest hindrance. Once culture has remained at a standstill for a couple of days. 14* 106 | 28 the ordinary saprophytic streptococeit). If several nitrogen sources be mixed together, giving a greater certainty of obtaining all the requisite building material for bacterium protein, then the specific qualities of the various sources in themselves will as a rule be effaced. The favourable effect of phosphoric acid upon the lactic acid fermentation is seen most distinctly by comparing the W, and W; less markedly in the experiments with casein peptone and different quantities of phosphoric acid, as the casein pepton in itself contains some phosphoric acid. As a rule, it does not seem that anything essential is gained by adding more than 2/5, A,gHPO,, and a single strain (Sbm. planlarum No. 1) even appears to be quickly satiated with phosphoric acid?). We have also tried some outside species of bacteria, with different sources of nitrogen. Table VIII shows that the propionic acid bacteria exibit a great partiality for yeast extract, whereas the omnivorous coli bacteria of course exhibit a far slighter sensibility in regard to the species of the nitrogen source (as also, by the way in respect of temperature and many other factors) than the true lac- tic acid bacteria. As regards the aerogenes bacteria, the acid formation can here no longer be taken as any measure of the vital activity, as these bacteria turn more of the sugar into gas the better they thrive, and under favourable conditions, the inoculated tubes become even less acid than the control tubes. Having investigated the question of which nitrogen sources the different species of bacteria prefer, it is then necessary to determine at what degree of concentration they should preferably be employed. As with sugars, the colour tone and price of the nutritive substrates are increased with increasing concentration, and it is therefore better to keep a little below the optimal limit than to exceed it. It is obvious, of course, that the amount of acid formed cannot be any thoroughly valid measure of the optimal concentration of nitrogen, as all organic nitrogenous nourishment acts as a buffer, and therefore a greater quantity of acid may thus possibly be formed — without exceeding the hydrogen ion con- centration detrimental to the various species of bacteria — the more nitrogenous nourish- ment is given. Nevertheless, we have also in the present instance kept to this measure, not only because it gives a numerical expression of the vital activity of the lactic acid bacteria, but also because further study has shown that lactic acid bacteria really grow most rapidly, and are able to ferment most sugars (i. e. become most abundantly sup- plied with enzymes) in those nutritive substrates in which they form most acid. For the experiments shown in Table IX, we used a 2% solution of grape sugar, with the necessary salts, and Witte peptone or casein peptone as source of nitrogen. Of Witte peptone, we used 4%, 2,5, 10 and 15%, answering to the quantities of nitrogen shown in the table. These are, in the case of the 10 and 15% Witte peptone broth, somewhat lower than they should be, as some ingredients of the Witte peptone are no longer fully soluble at these concentrations. As regards the casein peptone, we used partly the solution (with 1) In the case of these streptococci, yeast extract appears to be as good a source of nitrogen as casein peptone, judging merely from the quantity of acid formed; this is, however, due to the fact that the yeast extract has fully as great a ‘buffer action as the casein peptone, a feature which we shall refer to more fully later on. ?) As the experiment was repeated several times, the figures given are not due to accidental cir- cumstances. 29 107 1% N) obtained directly from digestion with pepsin, partly the same diluted to twice and four times the volume. The solutions were of course neutralised, and all contained the same quantity of nutritive salts. The table shows distinctly enough that more acid is formed with increasing quanti- ties of nitrogenous food. When the higher concentrations of nitrogen are reached, however, the increase does not always amount to anything worth mentioning, and in the case of the betacocci and betabacteria, the highest concentrations even appear to be detrimental. Despite the buffer action of the nitrogenous food, the quantity of acid cannot increase indefinitely with increasing quantity of nitrogenous nourishment, owing to the fact that, as van Dam has recently shown!), the lactic acid fermentation is not only checked by the hydrogen ions, but also by the lactate ions. It is also a well known fact that only the most powerful lactic acid formers are capable of fermenting all the milk sugar of the milk, even where chalk has been added, and the acid formed has been neutralised by constant shaking. These complications render it impossible to set up any hydrogen ion concentrations as the limit of fermentation for the different species of lactic acid bacteria. The better the buffer action of the nutritive substrate, the more the lactate ions will make their presence felt, and the final concentration of hydrogen ions will be lower in consequence. The lactate ions, however, are not nearly so dangerous to the life of the lactic acid bacteria as the hydrogen ions, and in cultivating lactic acid bacteria, therefore, care should always be taken to employ nutritive substrates with a good buffer action. It will be seen from Table IX, that the broth with 0.5°% N. in the form of casein pep- ton is a better nutritive substrate for lactic acid bacteria than broth with 0.7°% N in the form of WITTE peptone, and can in many cases even compare with broth having 1.35% N. in the form of Wrrre peptone. An das the casein peptone broth is also lighter in colour, than 5% or 10 % Witte peptone broth, and forms no deposit with acid, it will easily be understood that we preferred this for cultivation of lactic acid bacteria, and we use it with just 0.5% N. as any further increase of the concentration only exceptionally im- proves the action. The peculiar behaviour of the aerogenes bacteria towards sugar, to which we have referred in the course of the explanation of Table VIII, is also apparent from Table IX. The quantity of acid rises, it is true, when the quantity of Wirre peptone is increased from % to 2%, but when this point has been passed, we find that more and more of the acid formed is changed into gas, and as at the same time a certain protein decomposition 1) Ueber den Einfluss der Milchsäure auf der Milchsäuregärung (Biochemische Zeitschrift 1918, Bd. 87, p. 107). In this work, van Dam reproaches me with not having observed the influence of the hydrogen ion concentration upon the lactic acid bacteria, basing his accusation upon a brief statement made at a congress with regard to the present work, in which it was quite impossible to enter into any detailed explanation of the individual phenomena. I can, however, console my critic with the fact that Miss Jenny HEMPEL — three years before the date of van Dam’s paper — had kindly invest- igated, at the CARLSBERG Laboratory, the buffer action of the nutritive substrates which I was using, and found that it increased in the following order: WırrE peptone, casein peptone, and yeast extract. Yeast extract broth with 0.5 "Jo N has about the same buffer action as WITTE peptone broth with 1.35 "/0 N (i. e. with 10 °/o WITTE peptone). 108 30 Table IX a. 2. | ‘Nas casein peptone a Species of bacteria No. | | ” | 0,07 028 | 0,70 | 1,35 | 200 | 025 | 050 | 1,00 | i k i i i | Streptococcus laclis_...----- + 18 | 41 59 | 7,9 | 65 : ALES 6 |! 16 | 32 | 52 | 63 | 63 | » CE EST Tel 1,9 L 3.6 5,6 5,6 61 > eee ee 8 | 19 | 36 | 59 | 68 | 65 | XIV > ee eo. 9 | 19 | 38 | 56 | 74 | | » Bun AR ae a | 70 . BE. 14 | 20 | 38 | 58 | 63 7,0 . gå 22a 16 | 13 | 38 | 50 | 72 74 | DUT its +. ARE 17! 18 | 41] 50 | 79! 90! 47 | 63 | 68 ; feeium....... 8 | 18 | 38 | 50 | 61] 77] 45 | G2] 65 a abe RTE 14 | 16 | 36 | 54 | 63 | 831 52 | 68 | 68 x - n AT 171 | 26 | 56) 72 | 74 | 92 | | É a 8) 25) 50) 72) 7,7 Bu > glycerinaceus 1} 16 | 36 | 59 | 70 | 6,1 | | ; ? 3 | 16 | 40 | 50 22 | 92} 38 | 56 | 52 XXI > ; 4 | 14 | 38 | 56 | 68 | 65 | i , 6 | 16 | 34 | 54 | 68 | 61 | ' > liquefaciens 1 14 | 43 | 60 | 68 | 59 i XXII 4 : i 3 | 16 | 45 | 68 | 74 74 | bå Lær : mastitidis...... 2 | 11] 34] 45] 68 | 56 i . ee 3 | 13 | 34 | 45 | 61 | 50 | XVII x thermophilus 6 | 05 | 25 | 52 | 65 41 | > eremoris...... 1 | 11 | 29] 41 5,0 | 5,6 27 45 | | 3 ee 2114] 29| 47] 59 wi 36 | sıl 72 = : TRIER 10 | 11 | 29 | 41] 2 63 3 2 ee gt ar 11 07 2,6 43 61 65 | ; = Che I o | 36 | 50 | 81 | 83 | 41 56 | 77 > » sete 19} 18 | 20 | 34 | 52 47 | Betacoccus arabinosaceus 6 | 09 | 23 | 41 45 50 | : | É i 7) 09 | 25 | 43 | 59 54 XXV 7 > si a8 | 20 | 38 | 54 50 | ieee 9 | as | 23 | 41 | 50 52 | > bor +22 3 | 06 | 16 | 34 | 52 | 59 | ne | ’ en ufo | 15] 29) | 59 | a | i mé. tea 40 | OF | 25} 52) 65) 52] 36 | 36 | 41 > RAT 2 | 05 | 14] 27) 45 | 32} 36 56 | 4,1 takes place, the final result is, that the nutritive substrate becomes alkaline’), which, as already mentioned, has a detrimental effect upon the bacteria. Nutritive substrates with good buffer action are therefore not suited to continued cultivation of aerogenes bacteria. 1) On the basis of this interesting feature, which I pointed out in a lecture at the international dairy congress at Bern in 1912 (later printed in Zeitschrift fur Gaerungsphysiologie 1914, V, 10) Crank and Luss have, in the Journal of Biological Chemistry, 1917, Vol. XXX, p. 209, worked out an easy method for the separating of coli and aerogenes bacteria. 31 | 109 Table IX b. | °lo N as WITTE peptone °lo N as casein peptone Nis | Species of bacteria NO FE Te ae -——h 0.2 — | | | 0,07 | 0,28 | 0,70 | 1,35 | 2,00 | 0,25 | 0,50 | 1,00 | | ieee Goa Le xx vu { | Tetracoccus casei........... | 5 0,7 | 18 | 32 5,9 | 63 2,0 3,4 5,6 > 225: 7 0,7 1,6 2,9 4,3 4,3 1,6 2,9 4,5 XX VII f » liquefaciens . .. .. | 9 | 0,5 | 1,5 25 | 32 3,2 \ » Er; 10 | 05 | 07 | 14 | 23 2,3 XXVII > . PCR | 12 | 09 | 20 | 27 | 52 | 52 | 25 38 | 5,6 VIL » mycodermatus .. | 31 | 0,7 1,7 2,7 3,6 3,4 | Thermobacterium lactis..... Cine at rit | as Ma 1,6 ae > ETES CE Er | 0,9 | 1,0 | 6,1 12,2 » =, CRE | 11 | 13 | 4,1 7,7 | 7,7 10,8 XXVIII » helveticum | 12 | 0 | 0,7 14 | 74 | 115 0 38 7,4 XXVIII | » bulgaricum | 14 | 0 | 0.2 14 | 29 5,9 ae 54 7,4 "| Streplobacterium casei ..... | 4 | 28 4,5 8,3 | 9,7 10,6 > Ba | 5 | 29 | 55 | 70 | 86 | 12,4 » Re n- I aed4>| 631568 | 97 10,4 ERIK CR TR TE | 9 | 3 4 1,0 7,7 | 9,0 11,5 » “Pees | 11 | 18 | 63 | 7,2 | 133 | 124 7,2 12,8 : en lig | 23 | 5,5 | 61 | 119 11,3 » Oe — 321 36 | 83 | 99 | 122 13 » ree . 34 | 38 | 8,1 | 117 | 11,9 14,4 | » plantarum 3 | 0,7 5,4 TA 95 | 11,5 7,7 11.0 | = » aan! 4,1 | 50 |. 54 | 10,6 6,5 Gt 7,9 ee à - 3 u | 14] 32] 47 }104 | 11,90 | 74 | 74 | 74 | à » 3-118 | 63° )°63 168 |u5 I 79 7,9 | 79 | À > 15 | Lt | 37 1063 | 22 1117 6,8 it to | » » | 21 | 28 5,4 i 7; 9,9 12,8 | | Betabacterium breve........ ee 0 3,8 | 6,1 | 5,9 5,2 XXXI | : ps | 5) 16 | 56) Tt) 59 SE | » ee Ti 0 |.36 | 38 |.61 1,7 » >) See 10 | 0 1,4 25 2,9 05 2,0 4,1 4,1 Microbacterium lacticum.... | 2 | 0 02 | 14 | 37 3,6 So » » j re | 3 | 1,1 2,2 | 2,5 4,1 38 > ade at Wie | 23 fae as 3,6 IERRIT | » flavum ..... 8 | 18 29 | 5,4 | 83 6,5 2 | Bacterium coli A........... 1 | 18 | 29 | 36 | 54 5,2 | 2 eee es. 4 | 16 | 29 | 36 | 50 4,7 | » aérogenes ....... th Er} 2911 OG + 0,7 | » » 2 1,1 3,0 | 0,7 = > The same applies to the fluorescent bacteria, which, as typical water bacteria, are best without any over-abundance of nitrogenous food, with which they lose in the course of a few generations the power of forming colouring matter and then perish altogether. As already mentioned, the lactic acid bacteria can thrive when there is only a trace of the requisite carbon sources present (they grow, for instance, in agar without added sugar) and many streptococci, and even some few rod forms, exhibit a certain growth 110 32 tn sugarless broth, when they have good sources of nitrogen at their disposal). The lac- tic acid bacteria are far more difficult to satisfy in respect of nitrogenous nourishment, and it will be seen from Table IX, that several strains are unable to ferment grape sugar, if they have only 15% WITTE peptone available. Their demands in regard to nitrogenous food are further increased when dealing with sugars more difficult of fermentation, such as for instance inulin, which is often only affected when certain definite sources of nitrogen are present. Only when the nitrogenous nourishment is in all respects sufficient lactic acid bacteria are able to produce the numerous different enzymes (invertase, maltase, lactase, inulinase, etc.) which are required to decompose the di- and polysaccharides”). It is there- fore absolutely necessary to know the best sources of nitrogen for the different bacteria, before proceeding to investigate which sugars they are able to ferment at all. In the various tables for the different species of bacteria it will be seen how greatly the source of nitrogen may affect the fermentation of the sugar. We will especially draw attention to Streplococcus bovis Nos. 1 and 2 (Table XVIII), Betabaclerium breve No. 11 (Table XXXI), Microbaclerium laclicum No. 2 (Tab. XXXII) and the many strains of Streplobacterium plantarum (Table XXX), which are particularly susceptible as regards nitrogenous food. In all our earlier investigations with the lactic acid bacteria, WITTE pepton was our sole source of nitrogen, but even after we had found other and better sources, we continued nevertheless to employ WITTE peptone in addition, as it is only by studying the fermentation of sugar as well with good as with bad sources of nitrogen that it is possible to obtain a proper impression as to which sugars the various strains prefer. The aerogenes bac- teria are here, as in most other respects, found to take up a reverse position to the true lactic acid bacteria, as they will most easily render the nutritive substrate alkaline with the sugars which they find it most easy to ferment, and as a matter of fact, it is only with a slight quantity of nitrogenous nourishment that we can form any estimate as to which sugars they ferment, unless by measuring the quantitites of gas developed, which is the only rational method when dealing with these bacteria. The lactic acid bacteria use their carbonic food chiefly as a source of energy, and consequently throw off therefrom a quantity of breaking down products (fermentation products); their nitrogenous nourishment, on the other hand, is employed principally as building material, so that they do not need to give off any considerable amount of breaking down products. As a rule, no perceptible decomposition of nitrogenous matters lakes place at all, unless the bacteria are suffered to continue their vital activity for some length of time, and it can only be occasioned by gradual neutralisation of the acid as it is formed. Cultures in which it is required to study the decomposition of nitrogenous !) In this respect, however, there is a very essential difference between true and pseudo lactic acid bacteria, as the latter — at any rate under aerobic conditions — always thrive excellently in sugar-free peptone solutions. *) It is not unlikely that the entire question of vitamins, of which I have recently given a survey in “Naturens Verden” 1917, also has some connection with the enzyme production of the organisms. Jacoby for instance, has shown (Bichem. Zeitschr. 1918, Bd. 86, p. 329) that the formation of urease in bacteria capable of splitting up urea is greatly furthered by the presence of a certain amino-acid, to wit, leuein, 33 111 substances must therefore have chalk added, and this is the more necessary, since the proteolytic enzymes of lactic acid bacteria can only exceptionally be active in acid liquids. Only Streptococcus liquefaciens, and some few tetracocci (micrococci and sarcine) split up casein in sour milk, but they do so to a far greater degree when the acid is partly or entirely neutralised. » By way of measuring the degree to which the protein decomposition takes place in the milk, we have, as in previous works, determined the quantity of soluble nitrogen (SN) and decomposition nitrogen (DN) both expressed as % of the total nitrogen present. The former is simply the nitrogen in the filtrate, after the casein, albumin and globulin have been precipitated by boiling with a small quantity of acetic acid. As we work with sterilised milk, however, where the albumin and globulin are already coagulated, the boiling is generally unnecessary, and where the casein has been separated off by the lactic acid fermentation (the casein is precipitated despite the addition of the chalk) it is like- wise superfluous to add the acid, so thatas a rule, SN will mean the nitrogen in the filtrate of the culture. This filtrate is also used for determination of DN, by which is understood the nitrogen which is not precipitated by addition of sulphuric acid and phosphotungstic acid in certain definite proportions). As with the determinations of acid content, where we always subtracted the degree of acidity in the control tubes, so also here we subtract the SN and DN of the original milk (sterilised with chalk). What we want to accertain is, of course, the quantities formed by the bacteria. These quantities may quite well be negative, provided the organisms in question consume more than they give off, and DN may also be slightly in excess of SN, in cases where any of the original SN in the milk has also become broken down. In peptone broth, where all the nitrogen is found in a state of solution, the determination of SN will no longer be required; we can, however, on the other hand determine the amount of formol-titratable nitrogen according to S. P.L. S@RENSEN’S method?). It will nevertheless be necessary first to distill off the ammonia in the cultures with barium carbonate, and then remove the phosphoric acid with barium chloride and baryta. From the quantity of formol-titratable nitrogen found, we have throughout subtracted the amount of formol-titratable nitrogen in the non-inoculated broth, and by FN, therefore, we understand the formol-titratable nitrogen produced by the bacteria. In some cases, we have noted the first stage of the formol titration separately (i. e. prior to addition of formalin, from neutral point of litmus to faint red with phenolph- talein) and calculated the quantity of nitrogen corresponding thereto as a percentage of the total nitrogen content. According to HENRIQUES and GJALDBÆK*), this stage will as a rule be the less, the farther the proteins have been decomposed. The figures thus found should then be inversely proportional to DN, which is also to some extent found -to be the case (See Table X). In conformity with the fact that the true lactic acid bacteria are unable to live on ammonia salts or single amino acids, we find that they are also incapable of splitting up amino acids. They 1) 50 cm. of the milk filtrate — 30 cm.” sulphuric acid of 25 °/o + 20 cm.” phosphopungstic acid of 10 °/o are filled up to 250 cm? and left fo stand for 12 hours before filtration. For determination of nitrogen by KJELDAHL's method, 50 cm.” of the new filtrate is used, answering to 5 cm.” milk. *) Meddelelser fra Carlsberg Laboratoriet 1907, VII, p. 1. 3) Zeitschr. f. physiol. Chemie 1911, Bd. 75, p. 363. D. K. D. Vidensk. Selsk, Skr., naturvidensk. og mathem. Afd., 8. Række. V. 2. 15 112 34 Table Xa. | Milk WITTE peptone broth Table B ‘ Stage I in No Species of bacteria No. | sg ing ER ee lo of formol titration | | figure | Streptococcus lactis............. I" aa Hal 10,3 8,7 30 » "RY ee 3 | 146 6,6 9,2 4,0 28 > PTE ea Tol SON 07054 14,0 9,2 20 » SP trae es 8 ile 2,4 11,7 11,8 21 XIV ; és TR 9 OR EE 16,5 15,2 28 » RAR TES 10 51 + 0,3 16,1 12,3 22 » Ve ERE RAS 12 0 10,6 6,1 5,8 35 » EX A LEP 14 0,8 0,5 11,0 8,1 21 » DIOS DCR 1 5,0 1,6 14,1 8,5 26 | » ees ae 2 8,5 2,4 7,0 9,0 XV » D MT ER 10 11,4 1,8 9,4 6,5 » De Le RER 15 20,4 8,0 8,2 8,1 33 | » Br ME ESSEN 18 6,7 0,9 15,2 6,5 XVI » mastitidis ........ 3 2,8 19 iil 7,8 35 [XXIV » ipyogenes ske 2 No growth 33,6 192; 44 » fæcium. a a 8 4,1 0,6 15,7 83 | xx { > Be 144] 44 |r02 | 225 | 130 » glycerinaceus ..... 2 4,3 + 0,6 | 14,5 14,8 » fae ON PAPE 3 0,3 0 | ot 8,9 | 32 ou | » SNA EN. AE 4 3,4 0 19,6 es 23 » Dri gle Bact Er 6 0,8 0,8 7,2 109 | | Betacoccus arabinosaceus....... Wises [877 + 0,4 0,2 0,5 | XXV | » TAS ME eRe 5 2,8 06 | 1,0 38 | 45 > NP ee ae) 3,7 16 | 0 DT | » Davis FAT 33 0,5 05 | 0,5 3,1 XXVI » ER ERTEILEN LE | 34 12,2 3,2 17 1,6 \ » DL Coe at +. Pete | 36 0 0,9 6,1 4,5 40 xxvır$| Tetracoceus casei............+. 25 1,3 1,3 13,9 13,9 21 \ » Bo yk EN 6 0,5 0,9 29,6 20,3 | 15 { » liquefaciens ........ 9 75 12 35,6 LS XXVII; | » eo TIE SENE 10 71 16 25,1 10 21 | | » RL MBE FØRE 1 71 38 13,7 8,1 XX VII » Vink gor ae DA oe 12 48 16 | 11,8 10,2 25 cannot therefore, form more ammonia than is found as such in the pro- teins or peptones employed. As they generally only form 0—2% (in cultures from 6—9 months old 6% at the outside) NH, from the total amount of nitrogen, the quantity of ammonia formed is not shown in the tables. Table X shows examples of the decomposition of protein occasioned by lactic acid bacteria. The cultures had, as mentioned, all been treated with chalk, and had been left to stand for a month prior to investigation at 30° (only the culture of Thermobacterium lactis at 40°). Only the gelatin-liquefying species 35 113 Table X b. | Milk WITTE peptone broth Fur Species of bacteria No. | | | ae | SN | DN DNC UCP hore | | | | titration ei | | figure XXVIII | Thermobacterium lactis......... 4.1188 | 18,9 34,8 | 21,1 Streptobacterium casei.......... 2 9,9 10,7 15,6 | 14,9 25 > PRIS TUE SIN VAT | 19,4 15,3 12,3 22 » On 0 Sars Ae ee 5 | 128 | 9,8 21,0 | 15,2 | » RE COMTE 6 | 13,1 |. 287 lad |. 120 | » I ae Bee 7 | CAT 80 | 32,4 26,9 | 20) > Ne GENRE 10 | 113 | 145 19,0 15,8 » N fus 12 | 11,2 10,5 16,1 14,2 18 » SA ae ieee ee 13 | 148 10,4 29,5 18,5 » SEM Duge 14 | 9,8 10,2 14,7 12,3 XXIX 5 FAIR eee 16 | 6,7 85 | 186 | 174 20 » DENE NE N 18 6,4 7,9 189 | 123 19 » a... 19 | 126 15,9 VOM rg 19 » 17 LEUR | 22 | 13,6 17,5 217 | 15,0 » MER POLE | 24 | 10,7 13,8 14,0 | 14,1 » D GE 27 | 7,2 6,7 29,7 23,9 13 » ee ee 28 | 146 18,6 41,3 | 29,9 15 » NACRE ROIS 2 | 22,7 23,5 26,6 21,8 | 18 » DEN ewes | 33 | 18,7 22,0 18,8 14,6 » plantarum..... 1 | 0 0,1 8,0 4,9 30 » pS PIERRE 5 = 0,8 0,5 0,8 2,4 53 » ET}. 8 = 0,3 = 1,9 | 11,4 10,3 | ot » N ee 11 | 1,2 pil a 19 EX » J Ae FR es: | 3,9 1,0 8,8 3,5 30 » "SE 2 ces 14 | 2,9 2,6 8,5 5,6 » DE Na cate 15 ial or 0.6 | 3,2 4,7 44 » eth AAA 20 16,3 18:72 17312 33,9 » sahen“, 21 15,7 173 | 144 16,9 Betabaclerium breve. .......... 3 0,8 +05 | 19 1,1 SRE ee : 5 6,3 0,1 3,2 1,0 XXXI » ye Or EVER 6 | 2,4 0,6 2,1 1,8 ER her; 5,0 + 0,5 7,4 5,1 SIR RE ce Saha | 10 lé 3,8 0,3 2,8 | er 0,7 XXXII “ee lacticum ....... 511200 5,6 FRONT " XXXL » flavum,.. +... SEE. 0 0 STM je, 82 such as the tetracocci noted in the tables — show any strong solution of casein. Tetracoccus liquefaciens No. 11 even forms consider- ably more DN from casein than from Wırre peptone, whereas most of the other strains find it easier to break down peptone than the proteins of milk. The casein splitting cocci always form more SN than DN; the thermobacteria and streptobacteria on the other hand, if capable of splitting up casein, form as much SN as DN. Asit is chiefly monoamino- 15* 114 36 acids which are not precipitated by phosphotungstic acid, it is thus likely that the rod forms in question simply peel off the monoamino acids from the casein molecules. That this really is so, we have shown in the case of a single thermobacterium (Tbm. helveticum), as the nitrogenous components not precipitated with phosphotungstic acid could not be further split up by boiling with hydrochloric acid. As they showed the same formol titration figure before and after boiling, they must consist of free amino acids. It is somewhat different with the nitrogenous decomposition pro- ducts, formed by lactic acid bacteria from Witte peptone; they must evidently contain a quantity of complex compounds (polypeptides) not precipitated by phosphotungstic acid, as otherwise, DN could not in the great majority of cases be greater than FN. Disregarding the few species capable of liquefying gelatin, the proteolysis produ- ced by lactic acid bacteria is most active in old cultures with many dead cells. There is no doubt that the hydrolysis of proteins — like that of the sugars — is due to endoenzymes, and does not therefore spread in the nutritive substrate until the cells have become weakened, or even autolysed, It is the pro- teolytic enzymes of the lactic acid bacteria which occasion the ripening of cheese, and this process only reaches distinct development long after the sugar in the cheese has been fermented, when the lactic acid bacteria have ceased their activity, being either dead, or in a latent state. The fact that lactic acid bacteria only decompose pro- teins to amino acids, but do not decompose the latter any further, makes the maturing of cheese equal to the process of digestion, and not to that of putrefaction, which is particularly characterised by a breaking down of amino acids. We need not here go into the proteolytic qualities of the different species. They are least developed in the betacocci and betabacteria, but even among the betacocci, we may find a few strains, isolated from milk and dairy produce, able to split up casein to a perceptible degree. On the other hand, casein-splitting cocci may, when they have not for some time been cultivated in milk, lose their power of utilising casein as nitrogenous nourishment, and thus often the power to grow in milk at all, even where they are other- wise able to ferment milk sugar. The bacteria have, as mentioned, their own power of gradual adaptation to a new nitrogenous food, and great caution should therefore be observed in taking their proteolytic qualities as species character. Relation to Common Salt. In order to be perfectly sure as to the most favourable conditions of nourishment for a bacterium, it is necessary not only to know how they behave with the various carbon and nitrogen sources, but also with different nutritive salts. We have, however, already referred to the attitude of lactic acid bacteria towards the most important one of these, viz. potassium phosphate, and will here restrict our- selves to a brief mention of sodium chloride, not because this salt is necessary to their development, but because lactic acid bacteria are often found in substrates such as cheese, sour cabbage, not to speak of anchovy pickle, in which common salt is abundant. Table XIa and b shows the effect produced by greater or smaller quantities of sodium chloride upon the formation of acid. The casein peptone broth employed, which according 37 115 Table Xla. EE | Casein peptone broth with 2°, dextrose | d °» sodi i mg Species of bacteria ecco rumme aa = em | 0,5 | 2,5 5,5 10,5 15,5 | | Le ER a | = — . — 2 Zed = Streptococens lactis: | 3 4,3 4,2 | 1,6 0 0 » sm Ee Le. | 8 4,9 3,8 175) 0 0 ee | : ‘i ae ON a 21 3,6 0,9 0 0 » le aa CE a 16 4,3 4,6 1,1 0 0 » > sa N. N ET 53 | 28 0 0 » 5 1 MR Senet À ES POP PA hy 75 eat RC © 0,1 0 » pyogenes........... 10-105 Gk 2 ES oT tab 0 0 » eo 3 | 16 1,5 0 0 0 EIN | » See 7 3,5 2,8 0 0 0 » | BES Roe 9 2,1 2,2 0,6 | 0 0 XVI > mastitidis .......... 2 2,5 2,0 0,4 0 0 » NET dE 18 5,0 3,6 0 0 1720 XV { » De N; 20 4,5 3,8 0,8 0 0 XVII » thermophilus ....... 2 4,6 3,3 0 0 | 0 XVIII > LOC ee ee 1g dee: © 1,6 LOUE St SEX » inulinaceus ......... 6 3,6 1,8 0 0 0 N » glycerinaceus ....... 1 4,5 4,4 3,1 0,7 0 XXL » I 3 4,4 4,4 3,3 0,7 0 \ i is Ee | 6| 43 | 44 3,1 0,7 0 » liquefaciens......... 1 45 | 42 | 28 0,5 0 aa {| » NEL tai 5 4,5 4,0 3,1 0,5 0 | Betacoccus arabinosaceus ......... 1 3,4 2,8 1,3 0 0 | » id, APRES 11 | 4,5 40 | 235 0 XXV | » PCR ot Shorey. 12 4,5 4,2 2,8 0 0 | » ds "RCE 20 4,6 4,5 3,5 0 0 | » Jor Ne SR FRE 29 29 | 25 | 13 0 0 XXVI | » SE AE RE A 3,8 | 37 0,5 0 0 | > 2 eee 46 | 38 | 20 0 0 0 to the manner of preparation contained 44% NaCl, was given a further quantity of 2,5, 10 and 15% respectively of KAHLBAUM’s purest sodium chloride. As will be seen, only a very few of our bacteria are affected by the presence of 2.5% salt. Only Sc. cremoris, Sc. ther- mophilus, Sc. bovis, and Sc. inulinaceus are somewhat retarded in their development thereby. On the other hand, Sc. faecium, certain tetracocci (Nos. 1 and 3) and streptobac- teria (Sbm. casei Nos. 32 and 34 and Sbm. plantarum No. 3) grow better with 2.5% than with only 1% common salt. We find, however, that 5.5% salt is more or less harmful 116 38 Table XIb. Casein peptone broth with 2°/o dextrose A ù : eg Species of Pacers NG. and °/o sodium chloride | 0,5 2,5 5,5 10,5 15,5 | | | | Ter ACOCLUS a7 ee ere Re: Le 1 27 | 36 | 36 2,9 | 1,8 XXVII- » Sens oo8* -Seehs gees: = 2,9 3,2 2,9 2,3 1 6 | > =e TA SAN à | 15 2,3 23 2,3 1,6 0,9 » EN NE! 23 3,2 2,8 2,7 2,2 143 ‘ pyogenes aureus 13 3,1 2.1 2,5 1,8 0,7 » » albus 58 | 29 | 2,5 2,1 2,5 1,8 0,7 XXVIII Thermobacterium lactis .......... 1110 82 0,4 0,2 0,2 0 Streptobacterium casei............ 11 | 10,0 8,3 1,1 0 0 STEG » Dons. se some 13 | 6,7 6,4 | 0 0 0 | » RAR ER ro 2} 32 E68 1,3 5,4 0 0 » RE SR oe 34 95 11,3 4,1 0 0 | » plantarum ...... | 31 60 7,2 6,0 0 0 » Sl she: 6 | 64 5,2 3,2 0 0 » ‘ Sal ce 77 5,6 5,5 0 0 ; ART 15 1 7784 6,1 4,2 u RES Xxx » Foe ae 14 | 74 65 | 4,2 8% oe » So nents: 201 8:6 6,9 63 | 0 | 0 ET 30 | 92 B01 RC EN | a » ite Pure 32 | 9,5 8,2 | 4,5 0 | 0 » re VU 4 | 54 45 | 34 0 0 f _ Betabacterium breve ............. EE ae 3,4 0 0 | 0 XXXI | » tte SEE 10 | 79 590128 ae (FE) » ee 19414 50 44 | 0 0 | 0 XXXI » JonQUIN Gs ESS | 1322, 1722256 29, >| 0 0 0 2 f Microbacterium lacticum. ........ RS 2,5 25. | 794 0 0 Sr 3 oo) eee | 6 27 2,7 0,1 0 0 SKET » mesentericum..... aac 759 11 | 0,4 0 0 à » avn RCE 8 4,6 54 | 45 2,2 0 XXXII : , Pied ete ai 3-1 “ear. | ee 83 | O7 | 0 to all lactic acid bacteria, and 10.5% salt stops the growth of most of them. An exception is formed by the tetracocci (micrococci and sarcine) which can as a rule stand 15.5% of salt. The good bacteria found in herring pickle?) are also chiefly tetracocci. Te. No. 1, for instance, is from an anchovy pickle containing 15% NaCl. Possibly it may be the special shape of the tetracocci which makes them better able than other bacteria to stand the heavy osmotic pressure of strong salt solutions. Bc. bovis No. 46 however, is also a pronounced micrococcus, though it is not therefore able to endure greater quantities of salt. )) The bacteria which render herrings rancid are, according to investigations which we have car- ried out for "A/s Dansk Fiskekonservering” liquefying rod forms. eh eb rm" 39 117 IV. Attitude toward Different Temperatures. Important characteristic feature in a bacterium are the minimal, optimal and maximal temperatures for its vital activity, and the death temperature. The Minimal, Optimal and Maximal Temperatures for Vital Activity are determined simultaneously, by sowing the bacterium in a series of tubes with a good nutritive substrate, and placing them under observation at different temperatures. For the higher temperatures, thermostats with water jacket were employed; these could, thanks to the great heating capacity of the water, be regulated very accurately. For temperatures lower than that of the room, we used a Panum’s thermostat, heated with gas to 20° at one end, and cooled with ice at the other. Unfortunately, the temperature in the different com- partments varied from 1°—2°, according to the quantity of ice in the receptacle. All our strains were tested first at 5, 10, 15, 20, 25, 30, 35, 40, 45 and 50 degrees, and thereafter at certain intermediate temperatures if required, in order to determine the minimal and maximal temperatures more closely. Immediately after inoculation, the tubes must be heated in a water bath to the testing temperature, if this is over 40°, otherwise, some growth may take place before the air in the thermostat has communicated its temperature to the content of the tubes. At temperatures over 20°, the tubes were left to stand as usual for 14 days before titration; at 20° and under, on the other hand, they were left to stand for a month, as experience had shown that the maximum of acid is generally only reached after that time. When we are fortunate enough to strike the minimal or maximal temperature exactly, it will be possible to observe growth without acid formation, a state of things which is noted in Table XII by +. It should be pointed out that the optimal temperature for growth of lactic acid bac- teria cannot be determined by acid titration alone"), but only by daily observation, and where necessary by microscopic examination of the contents of the tubes. For the tem- perature at which the greatest quantity of acid is formed can lie some- what below the temperature at which liveliest growth takes place, owing to the fact that the acid is more destructive in its effects at the higher than at the lower temperature. This was very distinctly seen in the case of Sc. ihermophilus and the ther- mobacteria, which quite indisputably showed most rapid growth at 40°, but formed most acid at 30° and 35° respectively. And what is here said of the acid formation, applies in an even higher degree to the proteolytic action. I have previously shown that Tetracoc- cus liquefaciens (= Micrococcus casei liquefaciens)?) which grows most rapidly, and also ferments most sugar, at 30°, exhibits the strongest proteolysis at 20°, and BARTHEL has shown that streptococci’) whose optimal temperature is likewise about 30° can in the long run split up the casein in the milk most powerfully at indoor temperature. This question, however, is a more complicated one than that of the acid formation, since, as mentioned, the proteolysis occasioned by the lactic acid bacteria in the surrounding substrate is a pure enzyme action, which does not run parallel with the vital activity of the cells. 1) It can even less be determined — as many writers have done — by the time required for curdling milk, since milk will curdle with smaller quantities of acid as the temperature is increased. 2) 1. c. 1904, p. 32. 8) Meddelande Nr. 97 fran Centralanstalten för försöksväsendet pa jordbruksomradet 1914. 118 40 Table XII a. a ss | la las le lela 2 A 5 JADE Species of bacteria No.|82 |, || I ÿ ET] Pete | See No 3= il | She |] BE CO | one ao ne ee ae een Be DA ee Streptococcus lactis......... 4 | w 0 2,5 2,7 3,2 (4,0 4,1 13,2 12,5 | 0 0 0 » PRE Sat 6| » 010127! |3,2| 134186136132! [051010 0 > seo lay: 7|» 0 |0,9/2,3} 13,2! 13,513,513412,8/25| 0 0 0 » SUV APTE 8| » | 0 {1,1 2,9| [8,2]. 13,813,813,512,9 111010 0 XIV » DRE EH 9 | » 0|+123\" 129 3,8 |3,8 |3,6 |2,9 18/01/10 0 » SN pe ee 12 | » 0 0527| 1341 (3413513529 10,5! 0 | 0 0 » a EN 14 | » 01025] 1/32! 136183613634 |16| . 10,2] 0 | 0 » SA A AE 16 | » 01 0/29 2,7 3,2 3,2 |3,2 |3,2 2,3 |1,6| 0 0 » SHS ESS 17 | » 010123| |3,9| 13413,813813,4132| 0 0 0 » fecium....... 7 | M) 0 0,7 2,5 4,1 3,8 1,4 1,1| 0 | » BE Mure 8|W | 0,2} 125| sl 136136136127! 127! [2,7] 11,810 RR » RR. 14 | » 05| 118! 125! |3,313,5|3,5/3,3] 12,8} |2,0] 10,210 | » en eg 17 | » 0 7 33| 14315,6 15,6 14,71 12,9 2,9 10,8! 0 » iy Sek ah co 18 | » 0,5} (34) 138! 147155152171 13,1] |2,2|2,2) 0 XXIV » pyogenes...... 10 | € 6,9 7,7 8,8 7,9 6,8 15,6 XXIII » pyrite 7|w 0 2,5! 12,6] [2,9 13,7 |3,7 13,2 12,8 1,6| 9 0 XI » inulinaceus 5 LC 16| 1438) o lav] laa] ar 0,9| 1051010 » BODISL oy tae 1| M 0 0 0 0 |1,1 |2,7 |3,4 0 0 0 SVIN { » 37 aoe 5 | » 0 0 0 0 |3,4 15,2 15,4 0,5 0 | 0 » glycerinaceus 1 || W 0/0 [1,8 2,7 3,2 13,3 |3,6 |3,1 2,5 16/0 0 | ÿ » 2 | » 0 | 0 |L8 29 3,2 |3,2 [3,4 [3,2 2,5 1,6] 0 | 0 XXI » » 3| » 0 18| 1235| 134137188132] 12,7] 12,513,0| 0 | | » » 4 | » 01/10/23! |2,7| 134134138/36| [3,21 (22116|0 > » oly Peat ee 010116! |23] |8,0/3,3/8,5/2,9| al 22] 10,21 0 XXII J » liquefaciens... | 1) » 0/0 2,3} |2,8| 134134142140! [23] 11,410,9| 0 \ > » 14 gs 01023! 12,8] 138411451386! 132| 131| 10,1] 0 XVII » thermophilus . . 2! M 0 0 0 1,6 7,7 (7,5 72 1,1 0,5} 0 XXIV / » Pyogenes...... AVG | 0 0 0 27 3,8 0,5 0 0 \ » DEVICE 9 | » 0 0 0 4,5 5,2 2,7 0 0 XVI f » mastitidis..... 21 Ww 0 0 0 1,4 |2,6 |3,4 |2,6 |2,6 | + 0 0 Rei | » DATE eh 3| » 0 0 0 0,9 13,3 13,4 |2911,6| 0 0 ‚0 » cremoris ..... 1] » 0 | 0 0,7 14| [1,8 |2,2 |2,2]1,8| 0 | 0 0 0 | » » 2|» 0] 0 051 231 123,5 12,912,9 2,6 |2,3| 0 0 | 0 xv » > 11| » 00102! 127! |3,212,9|2,6|1,4| 0 | 0 0 0 » yes ss 18 | > 0 | 0 230 2,5 3,3 13,6 13,6 12,5 | O | 0 0 0 | > ARS 19/M|0105| 16,8| 72] |7,2|6,8|6,6 |5,2|1,6] 0 9 0 | | À » 0. | DU» 109 16,3 16,5 7,2 17,0 16,5 5,410 | 0 0 | 0 As it is hardly likely that the optimal temperature for the proteolytic enzymes should lie lower than that of the bacteria themselves?), the explanation must simply be that these enzymes are better preserved at a somewhat lower temperature. 1) M = milk. *) It is very common, on the other hand, to find that the optimal temperature for the enzymes both of microorganisms and plants lies far above the optimal temperature for vital action of the organisms in question, which merely shows that these enzymes are particularly resistant to the effect of heating, since all enzyme action increases, in reality, with the temperature. 41 . 119 Table XIIb. | lo | | x 4 | | | GE: | > || omnis mo |S | | | ag" lo o | ae, Species of bacteria |No.| 22 | T 7 7 MT LS Peau hae Weaker ep feed E No. | | = 04 Weis Ål æ |o |o Ja ON} oF |) 60 >: Ba > i iP u = | SAG | 2 Se Perera eet Le |. | | | 2 SÅ | Yen A | || LES ES A EN BT: 27 | | — | Belacoccus arabinosaceus.,. | 1a) W. 0 0 0,2 1,3| 1,7} 2,3} 1,6! | 0,7 | 0,5 010 » » | 4] » | 4 (0,7 2,0 12,2 1,9! 1,8) 1,8) 06101 0| | ¢ 0 XXV » » 5| » | 11,6 1,6 2,0 2,7| 2,0] 3,01 1401| 0| |0| |0o » » 6| » 1009| 120! 123| 126| 2,3) 2,31 1607| 0} | 0 | 0 » » | 8! » 1005| (20! 123| 12,6] 2,41 2,01 0,90,7| 0! | 0 0 | » » | 9! » lolo2! (20! 251 1236| 23 2,31 1,1091 o| | 0} 0 » Dons 1... .33 | » 0 0,27% 107 1,7| 1,6) 1,6! 0,70,1| 9 Poy? vO » Là: took fs: 35 | » 0 10,514| 1,8) {2,6} 2,5) 2,1] 1,1) 0 | 0 | 0 | 0 XXVI » SAS ES SERRE 37 | » | + 41,1 1,4 2,0 2,3 | 2,0) 1,8| 0,910 | O 1, One Mg] » bp TRIER 40 | » 0 | + 12,5) (29| 129} 2,9) 2,5) 20! +| 0] | O| |o » HS rere 42 |» Opa RL OMS IN ILES | 1,9) 1,6) 04/0: | 0, | 0} 0 Tetracoccus casei..... ..... 5 | » 0 + 0,2 1,6 1,9 2,0) 1,6 01 | 0 0 xxvit{ » ‘| TEST RER 6| » 0 + 0,6 1,7| 2,3| 23} 1,8/1,4| 0 | 0 0 » oe is ee Tee | 0 +| 0,7 1,6 | 1,9) 2,0! 1,6 0,11% 1:0 0 XXVII » liquefaciens..... 11 | » 0 | aE 0,7 1,8 | 1,8) 1,9) 1,80,2| 0 0 | 0 CURRENT DE 12 | » 0 0 | +105 1,4| 1,6| 2,2! 2,0 1,818| 0 | 0 XX VII { sy oe SENS 2%5| » 0! 1+10,610,7! 10,7] 09) 1,4} 0,71 0 | 0 o| lo X¥XVII » mycodermatus .. | 31 | » 0 0,5 0,5 1,1| 1,4) 1,5) 1311,1| 0 0 | O f Thermobacterium lactis .... 6 | M 0 0 0 0 |11,5113,1112,6 10,1 9,7 0,5| 0 XX VIII » win: 7|» 0 0 0 0 7,0 111.5} 10,8; | 9,910 \ » ra chk: 8| » 0 0 0 5,4 | 9,7 11,7) |11,7|: [113,0 XXVIII » helveticum | 12 | » 0 0 0 0 |16,0/27,4/28,1 21,6 19,1 7,91 0 XXVIII » bulgaricum | 14 || » 0 0 0 0 |10,6)14,2/13,3 12,6 10,8) /10,4/2,5 Belabacterium breve ....... 3 | w 0! Jo| lo2| [151 3,4! 3,8] 3635| 0| | 0 0 » 14 IR al» 0 0 05| 1,1} 2,6! 3,6| 3,4/3,3| 0| | 0 0 » TRES 5| > 0 0 | [0,7| 0,7 3,4] 220,5| 0| | 0 0 XXXI > ae: 6|» o! lol lo2| 10,7] 2,3 321 3206| o o| |o » a ERE 7 |» 0 | 0 0,2 0,5! 2,7| 3,6! 311,61 0| | 0| 0 » Ber ar: |10| » 0 0 0,2} [08| 0,9} 1,5! 3315| O| |o 0 | Microbacterium lactieum ... 2 »1) 0 0,2 | 1,4 11,6 19111100 ES ol 0 xxx | » > 3| » 0| |+ [06] 1,4! 1,9] 2,0) 07|0| 0} | 0 0 » i 6)», |0 0 + | 11,41 1,9) 2311801 0! | 0 0 > flavum..... 8 | »7)! 0 | 0 +! 0,2) 1,9] 2,0) 0710| 0 | 0 0 XXXII { ; rn. pl.» | {o} jo; |+| (05! 1,1) 1,41 08}0) 0} |o 0 In our earliest experiments with regard to influence of temperature upon the vital activity of lactic acid bacteria, where we were still unacquainted with the better sources of nitrogen, we employed exclusively WITTE peptone(W), or, where this proved altogether unsuitable, milk (M). In the later experiments, on the other hand, we used as a rule casein peptone (C). As, however, our first experiments included more intermediate temperatures 1) Later experiments with casein peptone as source of nitrogen did not change the minimum and maximum temperature. D. K. D. Videnskab. Selsk. Skr., naturvidensk. og mathem. Afd., 8 Række, V. 2. 16 120 42 Table XII c. | fF | | le, | o lo | ° | | | | Table oo » fe deel | S131 TSI NE TE es eee No Species of bacteria | No.) 2 | & „| „| | fof) olay Pe Fee LISlRlSIEISI | | SE Pei ela] sje] es] | | Ce] En a 1” | | FEINE Fi | | | | | | el ESz ET gl Streptobacterium casei...... 2| w 0; | 0) | 0} |05132|43j34/20|0| 10h |o | > TER TE | 41, | 10] | 005105 |23154154145 10,7| 1021010 > DSL 5] >| |o| |02) (14) |32/54159/59/36/0| [o| Lo > aor eee. | 91 >] | O} |02| [09 3,4|4,5/6,8/68/32/0] |0| lo 5 SNS. 110! >» | | O} 105! 105| 125136156/54111/0| |o 0 i NB u 14 o | 0! |+! 127155164158! 13,1] (04/0 |0 RRTX tec. [16h » | | O} joa! J1} 127138154145; [01] 0] 0 0 ‘ wes LÉ. mie RE 0| | Oo} 105 2,9| 4,0/5,6|5,6/5,2)0) | 0 0 RATS [2% | 2 | 0} fol! \05| |29154156/38/3410| |o| to : RE. 24/>| |ol lo1| los! |34/41/50/4,7/36|0| 10 0 ; RARES 138] >} jo) [02 11/ 14015115656) 1,1] 0 | 0 0 » Bee. 1:32 | » | 0} |0]+107| |2,6)5,6 81 74) 43| 1321010 : A LR 34 | > | O | 04, |2,5/5,1 8,1 7,4| 156| 15010101! plantarum. | 1! >» | |0| | 002/05) |23/56/61/47|11/0| |o| |o 5 > 5| >» | 0 0,2! |16123134117105/0| |o| Io 5 i 71>] | 0 0} 102! |59/82|73l61/36l0| | 0 0 : » 8} » | | o! (02) 11,61 |5,2|5,4| 5,8] 5,0] 4,2] 0 ol: lol » | 10 | >» | O0 |0 + 1815915,9136123|0| |o| |0] i > I1]> | | 0 0 +! |14/56142138/22|0 0} lo| à : 2!»| bol |o! Joe! Izzisıleılısiıslo| lol jo} » » 13 | » | O 0 |0,2/0,7| | 3,2] 7,0] 7,4) 2,4 01,010) | 0 > 14 | > | 0 0. A 1,1159/5,6147|19)0| |o| Jo $ 16] >| | 0 0 + 16122 32134118 0650| |o| lol! > IO ot 4224 POSEBOG 04; 14] 13,213,7|561654| I18slojo| |o | > tue STE 0 005105| |23/56/61/58/47/0| jo| | 0| | Bacterium coliA........... Y 1] » O| |20) |20| |23125/27|25 22| |20| 02) 0 > pot pie 2|>» | |o| 10,9! 12,0! |23/27/27/25 2,5| 241010 , seen 2 3| >» 0| |1,0 14| | 1,6/ 2,5 | 2,5 | 2,5 22| 201010! > SE ee RER Al » of 14 18| |2,3)2,3/2,6/23| 241 1231061 0 > paracoli......... 1 5» 0| | 0! l16l |24126/25120! {1,91 h:s8lolol paratyphi ....... | 6} > | | 0} {1,0 16! | 2,3) 2,5 | 2,5 | 23 20| 1,810! 0} aerogenes........ | 1] > La 1 2,7 | 2,7 | 2,7 (2,7) 1,6, 0,7 06 0 0 | » pentes: 2) > 0| | oO} Jı8] 235127127123 17, 116.050 > pat Ce 3| > 0 17| 118! |2,9/2,6)2,3) 1,6 15, 14/0101! SR A: 4| » 0| |0 18| | 2,0/2,6/2,7| 1,8 16, 1,5) 0 | 0 | than those after tests made later on, Table XII refers chiefly to the first experiments. The source of energy employed in the broth was always 2% grape sugar, and the figures given are as usual for the quantity of acid formed expressed in 9/44. It was found thatthe optimal temperature was independent of the source of nitrogen, whereas the upper, and particularly the lower limit for vital activity may be somewhat altered by improving the source of nitrogen. With casein peptone as source of nitrogen, for instance, we succeeded in get- 43 121 ting the great majority of the streptobacteria to form appreciable quantities of acid at 10°, despite the fact that the growth at this temperature is as a rule not perceptible until after the lapse of 14 days. The maxima] and minimal temperatures of the bac- teria are affected far more by the vitality of the bacteria themselves than by conditions of nourishment, and weakened strains therefore exhi- bit much steeper temperature curves than those whose vitality is unim- paired. Even though the attitude of the bacteria towards different temperatures may not be altogether constant, there is nevertheless hardly any other quality which better charac- terises the various species. It will be seen from Table XII a,b and c that Sc. fecium and the thermobacteria thrive well at 4714,—50°; some thermobacteria, indeed, at over 50°; that Sc. glycerinaceus, Sc. liquefaciens, and Sc. thermophilus aswell as the coli and aerogenes bacteria grow well at 45°, that Sc. bovis and the thermobacteria as a rule do not grow at ordinary indoor temperature, and that a few strains of Sc. cremoris and of the beta- cocci can grow already at 3°. As regards the optimal temperature, this lies, as already mentioned, in the case of Sc. thermophilus and the thermobacteria, at 40° or even a little higher; the same applies to Belabacterium longum (not shown in Table XII); for most pathogenic streptococci it is 35°—37°, whereas for all other lactic acid bacteria it is 30° or even lower. This last point cannot be too much emphasised, as many bacteriologists erroneously believe the optimal temperature of lactic acid bacteria to lie generally somewhere about blood heat, a temperature which on the contrary is detrimental to most of them. And in the fermentation test which is so important for the cheese maker, milk is placed just at this very critical temperature in order to determine whether good forms (i. e. true lactic acid bac- teria) or bad (i. e. the pseudo lactic acid bacteria) predominate therein. For at low tempe- ratures, the good bacteria, and at high temperatures the bad ones will too easily be able to get the upper hand (we have seen that the coli and aerogenes bacteria thrive well right up to 45°), even though they may by no means have been in the majority to begin with. The employment of different temperatures leads altogether to very active enrich- ment methods. Thus if milk be placed to stand at 45°—50°, then at first, Sc. fecium and Sc. thermophilus, more rarely Sc. glycerinaceus will gain the mastery, to be subjugated later by the thermobacteria, which are far more powerful acid formers. The Death Temperature. In order to determine the highest temperature to which a bacterium can be subjected without perishing, the capillary tube method is generally employed. It consists in drawing up a little broth culture of the particular bacterium into a series of capillary tubes, which are then fused up at the ends, and heated for a longer or shorter time in a water bath to different temperatures. Where the contents of the tubes exhibit no growth when sown out in a good nutritive substrate, the bacteria will have been killed. This method is subject to two sources of error. In the first place, we work with highly varying hydrogen ion concentration, according as more or Jess powerful acid formers — or possibly alkali formers — are being treated. And it is well known that the effect of heating upon the bacteria is greatly augmented the more the surrounding 16* 122 44 liquid diverges from the neutral point. In the second place, there is really no sharply marked limit at all which can be designated as the maximal temperature for the life of a bacterium, but experience has shown that most of the cells are killed off at a far lower temperature than is required to overcome the most resistant of the cells). It is therefore impossible to obtain a thorough insight into the question without investigating what takes place at each of the temperatures tried. In order to obtain accurate results, it will be necessary first of all to dilute the starting material before heating, to such a degree that its own concentration of hydrogen ions cannot produce any effect; secondly, to note the number of the cells sown out which survive the different temperatures. Both can easily be done by sowing out a couple of drops of the culture to be tested in agar tubes, and heating in the same. The surviving germs will grow out into colonies in the usual way, and can then be counted. We used Brrrrr tubes for this purpose, with a deep layer of casein peptone dextrose agar. It will of course be necessary to see that the agar has, prior to sowing out, reached the temperature of the water bath, and the germs must be distributed by rolling the tube to and fro in a vertical position, so that they do not settle higher in the tubes than the level of the water outside. When the heating is completed, the germs are thoroughly distributed by reversing the tube once or twice; the tube is then cooled as rapidly as possible, in cold water. When deal- ing with aerobic organisms — such as for instance the tetracocci — the agar is poured off into a petri dish after heating. Table XIII a and b shows the results obtained with some of our strains after heating for a quarter of an hour to 60°, 65°, 70°, 75°, 80° and 85° degrees. A temperature of 60° suffices to kill off the pathogenic species such as Sc. mastitidis and Sc. pyogenes. Sc. pyogenes No. 10 is an exception; this bacterium is, however, as we shall see later on, not a true Sc. pyogenes, but a pathogenic variety of Sc. fecium. At 65°, the betacocci are killed, and at 70°, the commonest lactic acid bacteria of milk, Sc. lactis and Sc. cremoris. Most of the cells of these will, however, have perished already at lower temperatures. Taking for instance Sc. lactis No. 2, we find from the table that only 4/49) % of its cells have been able to endure heating to 60°. A greater power of resistance is exhibited by the remaining lactic acid bacteria, of which some few cells can stand heating to 70°—75°. The most re- sistant species is Microbacterium lacticum, which is not always killed off entirely even at 85°. In milk pasteurised at low temperatures, therefore, we encounter chiefly this and Sc. thermophilus?), besides certain tetracocci (micrococci); somewhat Jess frequently Sc. fecium and Sc. glycerinaceus. Even though these streptococci only exceptionally survive heating to 75°, they have yet comparatively many cells which can stand 65°, andin practice, it is of far greater importance how the majority of the cells behave than what the most resistant individuals can stand, as these few cells will in any case. be unable to make their influence felt in the natural competition. 1) And even here we are not dealing with spore-formers, where the question is far more compli- cated. ?) In true low-pasteurised milk (heated only to 63°) Mcm. lacticum is, however, rare, at it is here unable to compete with the heat-resisting cocci. In milk heated to a somewhat higher temperature, on the other hand, we may find it almost as a pure culture. 45 123 Table XIITa. ‘a | | Millions of | | bacteria pr. | Amount of bacteria pr. cm? after '/, hour's heating to: Species of bacteria | No.| cm. in the — — ————— ————— ——— | | not heated | 60? 65° 70° 75° 80° | 85° | culture | | | T ae — { Streptococcus thermophitus. . 2 | 40 abundant) abundant |abundant 40 0 0 » » JR 5 || 12 » » 100 0 0 0 » fecium....... 6 | 46 » > | 2500 | 0 0 0 | » N, > 7 | 100 » » | 10000 | 40 0 0 » ~~ ae 8 | 50 > 30000 60 | 0 0 | 0 » “(a ee 12 133 » abundant | 340 80 0 0 » LEE 14 55 » 28000 | 1800 0 | 0 | 0 » Sy ERTL; : 17 | 50 » 40000 | 5000 0 | 0 0 » “a 118) 160 » 39000 11700 0 du, Ae » Pyogenes...... | 10 40 » abundant | 20 0 | 0 0 » glycerinaceus.. | 1 | not counted » » | 1000 20 0 | 0 » » 2 | » » less than at 60° | 400 0 0 | 0 » » | 4 50 » 20000 | 0 | 0 0 | 0 » » | ee: an à 7500 | 740 | 200 | 0 0 > > Me er orn los 15000 Och Spies 0 { » liquefaciens ... 1 112 » 3000 | 0 | 02] 0 0 » » ie 5 20 » less than at 60° | 360 0 0 | 0 » Eee MW 21 90 9000 200 | ei a ER 0 > Reeth es. 3| 160 200 0 | CS Be EEE 0 » Meme. Soc: Pit 40 2400 120 | A D 0 0 > ST 9 100 9000 20 | 0 on F0 0 > = ae 12) 150 | 0 0 | CA ae AN ae 0 | » BR... | 16 70 500 | 60 0 Ba 0 » =. 17 64 2000 20 0 | 01 0 0 » cremoris ...... 1 14 20000 200 0 | OPA 0 0 » D. 21 36 | 2500 1200 VA IS a adie 0 » TA LEP u | 16 2000 0 6} 0 0 0 » ls Eee | 18 | 8 800 20 0, | Oo | 0 | 0 » RU. [191 18 900 |. 0 Oe fe 0 0 > Rue... | 0 | 2 2000 2000 0 0 0 0 > mastitidis ..... a | 98 0 0 a da 0 | 0 XVI { i NÉ ::::. 2 | 2,5 0 0 o| o 0 0 > pyogenes ...... | 31 22 0 0 0 0 0 0 xxiv J > lie... 6 | Las, 0 0 0 0 0 0 \ > 2. mene i | 0 0 0 0 0 0 As a rule, those lactic acid bacteria which grow at the highest tem- peratures can also stand the highest degree of heating). 1) An exception is apparently formed by the highly heat-resisting Mbm. lacticum, which according to Table XII, does not grow at over 35°. As a feces bacterium, however, it must in a state of nature be able to grow at any rate at blood heat, and the result of our experiment must thus doubtless be due to the fact that the sources of nitrogen employed did not altogether satisfy this bacterium, which is very particular in this respect. 124 46 Table XIII b. | | Millions of | Amount of bacteria pr. cm? after 1/4 hour's heating to: | | bact Table Species of bacteria No. | u =: aa Ne. | not heated | 69° 659 | 70° 75° 80° 85° | culture I : | | | ( Betacoccus arabinosaceus … | 8! 105 6000 0 | 0 0 0 0 XXV » » | 11 100 6000 0 0 0 0 0 \ » » Sees op be 140 200 0 0 0 0 0 a » bovis. AB 33 120 0 0 0 0 0 0 nn | > à + Se idee 42 | 100 5000 0 | 0 0 0 Tetracoccus casei........... 5 | not counted abundant abundant | 60 0 0 0 xxvrr | » PRINT ER 6 | » | » » 900 | 200 0 0 » SAR Ac 7 | » > |» |less than at 65° | 1700 0 0 > . . 9 » i XXVII { > liquefaciens ...… A | å | ? | ig. = | s < 4 » CIE, PRE ER » XXVII | » mycodermatus... | 31 6,5 » 40 | 20 0 0 0 | Microbacterium lacticum. | 3 31 > abundant, abundant abundant 20 0 Nr » » 4 | 64 | » » | » | » 1200 0 XXX | » » 5 | 23 » | » | » » less than at 75° | 20 » » ENTRE | 6 | 7 » » | » 710 200 0 XX | » mesentericum | T| 0,5 » 1400 | 0 0 0 Ors | » flavum...... | 8| 1,6 ; abundant 17000 20 0 0 a Ü » op tee 9 0,5. | » > | 20000 5000 0 0 | Streptobacterium casei...... 2 58 | » 600 60 0 0 0 | » AE 4 | 40 | hers abundant 80 0 0 0 | » AT | 8 15 Fee 160 180 0 0 0 = » WE RE RR 11 | 21 | » abundant | 300 0 0 0 = | » “Mere BE 20 » 0 | 0 0 0 0 | » en | 23 | 21 > 5200 0 0 0 0. | » D. ities Bed | 24 | 58 | » 40 0 | 0 0 0 | > "eae | 34 | 47 | > 0 0 0 0 0 | » plantarum | 5 8 | » 100 200 0 0 | 0 | » | 6| 10 » 200 : | 120 0 0 | 0 XXX 3 | | » » | 14 | 8 | 320 0 | 0 0 0 0 AN » » | 30 16 | 9240 0 | 0 0 0 0 ( Betabacterium breve........ | 3 10 dan abundant 140 0 0 0 SET » Meee | OE | 6 | not counted » 120 80 9 0 0 | : Fante RS | 10 ig: COTES 1500 | 300 0 0 0 XXXI » longum...... | 32 | 15 | » 680 | 660 0 0 | 0 V. Other important Features. The expert in fermentation physiology should, in the identification of micro-organisms, lay most stress upon the phenomena of fermentation; the medical bacteriologist, on the other hand, will attach more importance to questions of agglutination and immunity. The latter are of but slight interest to us here, but an investigation into the agglutination of the lactic acid bacteria would doubtless in many cases have supported the results of our research. BARTHEL, for instance, has succeeded in showing, with regard to the ther- mobacteria, that the species established by us, Tbm. lactis, Tbm. helveticum and Tbm. bul- 47 125 garicum, have each its own manner of agglutination'). We were unfortunately debarred from making experiments in this direction, as our laboratory is not equipped for experi- mental work with animals, and with the great quantity of strains here concerned, it was likewise impossible to carry out the work as guests at any other laboratory. The results obtained by such agglutination experiments are, however, not always of the same interest from the systematic point of view, since, as will appear from M. CHRISTIANSEN’s work on the bacteria of the typhus-coli group?) the character of the agglutination need not always cover definite morphological, cultural or biological qualities. And this agrees excellently with the fact that, as we have mentioned, the bacteria are able to adapt themselves to new proteins. Among the most variable proteolytic qualities in bacteria are the hemolytic, and, as we know, everything connected with pathogenity at any rate for the pathogenic bacteria coming into our sphere of work (certain strepto- and micrococci) is among the qualities soonest lost on cultivation in artificial substrates. On the other hand, the researches of C. O. JENSEN render it likely that non- virulent coli bacteria — and in analogy therewith, also other non-virulent bacteria — can under certain conditions become pathogenic. Consequently, the pathogenic quali- ties cannot be utilised at all as species character, but merely serve to indicate whether we are dealing with a pathogenic variety. In LEHMANN and NEuMANN’s »Bakteriologische Diagnostik” (5. Edition 1912) taurocholate of sodium is used to separate off Sc. pyogenes from the remaining strepto- cocci. If 5—10% of this salt be added to broth cultures 24 hours old, of the various strepto- cocci, then their cells should dissolve in the course of some few minutes, save in the case of cells of Sc. pyogenes. This seems most mysterious to begin with, and involves the unfor- tunate conclusion that no streptococci other than Sc. pyogenes could live in the intestinal canal of animals, or at any rate, in that of carnivores. We on our part have never succeeded in observing the slightest clearing in broth cultures of any of our strains on addition of taurocholate of sodium, far less any real dissolution of the cells perceptible under the microscope’). Another reaction largely used for identification of streptococci is that based on their various power of reducing colouring matter. For this purpose, milk stained a pale blue with litmus’) is chiefly used. With like quantities sown out, and at like tempera- tures (which should preferably be somewhere near the optimal temperature for the bac- teria in question), the power of reduction will be proportional to the time taken in decolo- 1) Meddelande Nr. 68 fran Centralanstalten för försöksväsendet pa jordbruksomrädet. 1912. Tbm. Joghurt, which is otherwise nearest to Tbm. helveticum, agglutinates, however, together with Tbm. bulgaricum. 2) Det kgl. danske Videnskabernes Selskabs Skrifter, naturvidenskab. og matematisk Afd. 1916, 8. Rekke, I, 3. 3) We made our experiments with a preparation from Merck, and as this proved to contain at least as much sodium glycocholate as sodium taurocholate, we have ourselves prepared pure tauro- cholate of sodium from gall, by precipitating the sodium glycocholate from the mixture of gallic acids, with sugar of lead, and then removing the lead by means of sulphuretted hydrogen. Even the pure salt, however, gave no better result. It is only the cells of Sc. lanceolatus and Sc. mucosus, which are dissolved by taurocholate of sodium. 4) As milk, like other sugar-containing liquids, decolorises litmus during sterilisation, the litmus tincture must be sterilised separately, and dropped into each tube with a sterile pipette. 126 48 risation. Obviously, species which grow slowly in milk will also decolorise slowly. Se. pyogenes and most of the betacocci do not decolorise at all. Both spherical and rod-shaped lactic acid bacteria can as a rule decolorise litmus milks more rapidly than they coagulate it, but there are also species — Sc. thermophilus, for instance — which do not decolorise the milk until long after it has coagulated. As the power of reduction is in the highest degree dependent upon the vitality of the bacteria at the moment, it is of very little value as a species character. As has been shown by the present writer”) "and by BEISERINCK?), the true lactic- acid bacteria, in contrast to most other bacteria, are totally lacking in catalase, and the fact that broth cultures or surface colonies of the lactic acid bacteria do not develop oxygen with peroxide of hydrogen thus furnishes the principal test reac- tion for these bacteria. The tetracocci, however, are exceptions, which as a matter of fact do split up hydrogen peroxide to a very marked degree; the same applies to the microbacteria, which also in other respects behave somewhat differently. Of the lactic acid bacteria, only some of those which do split up hydrogen peroxide are capable of reducing nitrate to nitrite, but when dealing with such, this feature should always be tested. The easiest method is to cultivate the bacteria in dextrose broth with 2% KNO,; if nitrite is formed, then a small ladleful will form a dark blue spot in a mixture of zinc-iodine starch and sulphuric acid. With the knowledge obtained, through the investigations here described, about the lactic acid bacteria, their further determination will not as a rule occasion any serious difficulty. We always commenced, of course, with a microscopic examination of the isolated acid formers, in order to ascertain whether they were GRAM positive, and whether they were spherical or rod-shaped, but this is as far as it is possible to get by morphological investigation in the first instance. Not till we have learned the biological qualities of the strains, and grouped them accordingly, can we begin to consider the question of whether certain related strains may have certain morphological qualities in common, and this will also frequently be found — though by no means always — to be the case. The morpho- logical differences between the different species of spherical forms, or between those of the rod forms, or even, indeed, between their various genera, are so slight that they cannot be determined until we know what strains are related together, as only the total impres- sion of a great number of related strains can give a general idea of any value. After the Gram test, the next thing to try is the reaction with peroxide of hydrogen, whereby the tetracocci and most of the microbacterium are separated off. The attitude of the strains towards different temperatures is then noted, as also towards different sugars, and different sources of nitrogen. The acids formed, and also any splitting up of casein, are likewise further studied. ’) Det kgl. danske Videnskabernes Selskabs Forhandlinger 1906, Nr. 5. In this work it is proved that also the butyric acid bacteria lack catalase. *) Archives Néerlandaise des Sciences Exactes et Naturelles 1907, Serie II, Tome XIII, p. 357. % a 49 127 None of these stages in the investigation can be dispensed with, and none of them can be taken as of decisive importance by itself, not even the reaction with a long series of different sugars, though this does give such a valuable insight into the biology of the bac- teria. In the first place, it is necessary of course to know whether theacid formed is lactic acid at all, and in the second place, it is possible, as we shall see later on, for strains undoubtedly belonging to the same species to react very differently towards certain sugars. We cannot therefore lay too much stress upon the importance of avoiding the onesided classification according to reaction with sugars, which has been largely practised in particular by American writers. Only by taking equal note of all qualities can we arrive at a natural bacteria system. Variability of Qualities. Among the particularly variable qualities of bacteria are colour formation, and to some extent also the size and appearance of the surface growth; points which we shall deal with more closely when discussing the tetracocci. In this connection it will suffice to mention that comparatively anaerobic propionic acid bacteria transplanted from agar have in the course of years become slimy, and developed a marked surface growth, thus resembling in cultural respects the aerogenes bacteria. As mentioned, the bacteria can, under particular circumstances, accustom them- selves to new sources of nitrogen, but the reverse process — the loss of power to utilise a given source of nitrogen — is far more likely to occur, and on the whole, the varia- tion which we have otherwise encountered in the lactic acid bacteria, can be explained as a further development of already existent tenden- cies, or more frequently as the results of weakness or degeneration. We have never, for instance, found any of our strains acquiring the power to ferment a new sugar, though we have occasionally found that its power of fermenting a certain sugar, originally but slight, may be increased, and we have often noticed that it has lost the power of fermenting certain sugars at all, those being chiefly those which it had always previously found difficult to utilise. As an indication of weakness also, we should regard the point already mentioned, that strains forming both dextre- and lævo-lactic acid may lose the power of forming that of the two which it formed in the lesser quantity, and that strains which besides lactic acid also formed abundant quantities of by-products lose altogether or in part the power of so doing, by which they will of course be utilising their source of energy to a less complete degree. A highly variable quality is the power of forming slime in milk, but this is, as men- ‘tioned, only due to further development of a tendency possessed by all lactic acid bac- teria in a youthful state. It is interesting to note how closely-related strains are inclined to vary in the same way; or may, indeed, when cultivated under like conditions, often be seen to vary die off — at exactly the same time. All these apparent variations, therefore, are in reality no hindrance at all to species determination; on the contrary, the manner in which a bacterium is inclined to vary is often one of its most characteristic qualities. D. K. D. Vidensk. Selsk, Skr., naturvidensk, og mathem. Afd, 8. Række, V. 2. 17 and 128 50 The feature by which the alteration in vitality of lactic acid bacteria may be most directly observed, is the quantity of acid formed from a given sugar under uniform con- ditions, and even more so, the rate at which souring takes place; this can, in the case of milk cultures at like temperatures, be measured by the rate of coagulation. As the rate of souring (in contrast to the degree of acidity in the cultures after 14 days) is influenced to a high degree by the quantity sown out, we always inoculated the 10 cm? of milk employed with the same platinum loop, from a previous (just curdled) milk culture. Where greater fluctuations were observed in these respects, the extreme limits are noted in the tables. The highest amount of acid is shown first in the case of sugars whose fermen- tation has declined in course of years, and last in the case of sugars whose fermentation has grown stronger. If the lactic acid bacteria are to be of practical use, and keep down the detrimental bacteria, then it is primarily necessary that they shall sour rapidly and strongly, and this is just where laboratory cultures are often liable to fall short. If their acidulating power has once declined, then it is as a rule very difficult to restore it completely, even when they are transferred daily from milk to milk. Surprising results may, however, often be obtained in this respect by using a larger quantity of inoculating material. We have had strains of Streptococcus cremoris, for instance, used for the souring of cream in making butter, which could not be brought to produce more than 4.7°/,, lactic acid in twenty-four hours at 25°, however frequently we might transfer them with the platinum loop to a new 10 cm? of milk. When, on the other hand, we proceeded to a daily inoculation of 10 cm? culture in 200 cm? milk, we could then, after the lapse of two or three weeks, obtain in 20 hours the 7 °/o, lactic acid required for maintenance of activity in the pasteurised, though not therefore by any means germ-free, dairy cream. Next to unsuitable com- position of the nutritive substrate and preservation at too high tempera- tures, the slight quantity of inoculating material generally used in la- boratories is the main cause of the frequent degeneration in laboratory cultures). That an abundant quantity of inoculating material should give such favourable re- sults is due to the fact that we then are more sure of transferring some of the strongest indi- -viduals in the culture. If a pure culture be spread on agar or gelatin, and a series of new strains isolated therefrom, then the latter will never behave altogether alike towards the different sugars, and all will as a rule—untilthey have been repeatedly transferred forsome time — form a somewhat smaller quantity of acid than the original culture, as it would be a very fortunate chance to obtain on spreading, some of the most powerful individuals which stamp the culture as a whole. As a matter of fact, the majority of indivi- duals in a bacteria culture are weakened, and have but slight power of resistance; this was very distinctly apparent from our heating experiments. This explains why pure cultures are never at first so powerfully effective in practice as the original culture previously employed, and shows, that it is by no means a matter of indif- ference whether a person becomes infected by a few cells of a disease bacteria or a great number of the same. ') In the case of stab cultures of comparatively anaerobic lactic acid bacteria, as for instance the thermobacteria, it will be necessary to inoculate from the bottom of the stab, as the cells situated nearer to the surface will always be greatly weakened by the oxygen in the air. Description of Species, and Systematism. In discussing the separate lactic acid bacteria, it would be most natural to commence, as I have done in my “Dairy Bacteriology’, with the rod forms, which are as a rule the strongest acid formers, proceeding then by way of the streptococci, which are all typical lactic acid bacteria, to the micrococci, where we find all possible transition forms between acid-forming and non-acid-forming bacteria. For practical reasons, however, I prefer in the present work to commence with the streptococci, as it seems easier here, in most cases, to define the separate species than is the case with the two other groups. Streptococci. By streptococci we understand, as is generally known, spherical bacteria dividing as a rule in one direction only. Distinction is made between the proper streptoccocus type, and the diplococcus type, according as the cells after division are inclined to remain hanging together in long chains, or to fall apart rapidly. As a rule, the streptococci stretch before division, so that the cells are then oval. After division has taken place, the daughter cells are often egg-shaped, the pointed ends turning outwards. Where the growth is lively, several of the long-chained forms will not have time to stretch before division, but form disc-like segments. Again, the double hemisphere form, with flat surface at the break, so typical among the micrococci, is one which we have encountered in some species (Strep- lococcus fæcium, Streptococcus liquefaciens and Belacoccus bovis). In broth, the long-chained strains form flakes which easily settle, so that the liquor above them rapidly clears, whereas the short-chained strains remain suspended for a long time, so that the liquid takes longer to clear. These morhological differences have proved constant for all our strains of streptococci throughout the years during which we have had them under observation, and we cannot therefore refrain from considering them of some value as species characters, though it must be admitted that most strains in a weakened state form shorter and in particular far thinner chains than they did when at their full vitality. Here as with all other qualities in bacteria, it will be necessary to take into consideration the effect of temperature and of the nutritive substrate. Broth, for instance, increases the tendency to chain formation, whereas milk produces a reverse effect, and we have therefore only reckoned strains which also in milk grow in long chains as typical chain forms. In agar streak, and on gelatin plates, both types can produce rod forms, or other divergent forms. As regards 177 130 52 the influence of temperature, we generally find the most pronounced flake formation in broth near the maximum temperature. According to their biological qualities, the streptococci fall into two main groups. The one forms, from all carbohydrates, pure dextro-lactic acid, with only a trace of by- products, whereas the other forms levo-lactic acid, besides appreciable quantities of other fermentation products. The former group is by far the larger, and comprises not only the principal streptococci of milk, but also the pathogenic forms known under the collective term of Streptococcus pyogenes. They split up the sugars simply according to the formula C,H 720g = 2C,;H,03, and the quantities of by-products, carbonic acid and acetic acid (acetic acid and propionic acid generally in the proportion of 20:1), which are formed, are so insignificant that they do not practically speaking diminish the theoretical yield of lactic acid. To the other group of streptococci belong the bacterium of sour cabbage, Streptococcus brassicæ, and the slime former so well known in sugar manufacture, Strep- lococcus mesenteroides, which forms slime from cane sugar and mannite from lævulose, and which can develop a considerable quantity of gas. With regard to nitrogenous nourishment also, there is a difference between the two groups, the former preferring casein peptone and the other yeast extract. The first group, again, generally prefers low sugar concentrations, the second high (5%—10% sugar). ; These two groups differ altogether so widely in biological respects that they must be regarded as two distinct genera. And I therefore designated them originally as Dextrococ- cus and Lævococcus, but relinquished these names afterwards, on finding that some few bacteria belonging to the genus Lævococcus formed inactive lactic acid. For the first group, it would be tempting to employ the name Lactococcus, suggested by BEIJERINCK, if it were not for the fact that so many pathogenic bacteria were allied thereto. I have there- fore preferred simply to reserve the generic name Streptococcus for the first group. The unavoidable alteration of names is thus reduced to the least possible; the only thing is, that we are obliged for the future to give the term a somewhat more restricted meaning, and only understand thereby such streptococci as form dextro-lactic acid. The second group, being chiefly met with on souring vegetable matter and particularly on beets, L have accordingly given the generic name of Betacoccus (Beta — beet). Genus: Streptococcus. (Abbr. Sc.). This genus I have divided into the following well-characterised species: Streptococcus lactis, Sc. cremoris, Sc. maslitidis, Sc. thermophilus, Sc. bovis, Sc. inulinaceus, Sc. fecium, Sc. glycerinaceus and Sc. liquefaciens. In addition, there are also several strains — including pathogenic — which are allied to these, and cannot exactly be placed under the mentioned species, but we have not encountered them often enough to ascertain their entire range of variation, and I have not therefore ventured to establish them as separate species in themselves. Streptococcus lactis (Table XIV). By this we understand the diplococcus which predominates in sour milk, and which is known in literature under the names of Baclerium lactis (Lister 1878), Streplococcus acidi lactici (GROTENFELD 1879), Bacterium lactis acidi (LEICHMANN 1894) Baclerium Güntheri (LEHMANN & NEUMANN) and Sireptococcus lacticus 53 131 Table XIV. | RE ai TE a = w | w | | | Milk Streptococeus | 35 | .£ (lær 8lels 22181818 eje/2|.lele BER. No eet: ss|5]2 2813 33 E [2151312 | 2) 5/2 5je2]s, on of WR Ez »#|1:15181%2 8151815 82 & 3 5 5|23|32, Total N isolated from: |33 | = ME 235 R 2 85338 sl< 4 à) 35255 Tota ie “2°” a |) Sr "R2 4 tt 2,5 aa ae ‚100 4/4,0/3,8 2,8 10,113 14 | Sourmilk 4 180] 0 18.4/4,013,8'2,8 10,113,6 | | 0! 0 4,5 17,0/7,7/7,2 6,3 0,7/6,817,0 Kefir 1 0| 0| 0 32363629] 0 542,9 13,9) 0! 0! 0 383,915” 2,5|| 0 13,2) Sourmilk 2 2 ent 2,0 4 4 0 Lo 0 ee 0,5163 +6 0,2| 0| 0 2,5 /4,1/4,1/4,2 2,8 0,1/4,1/3,4) 0! | ve | | | - 0 13,6 |7,9/8,1/8,1 6,1 /0,517,45,20,2] 0 4,7.0,516,5 A Na ae 1) Sourmilk means milk, curdled spontaneously with acid. 132 54 (Kruse). As will be seen, this bacterium has been regarded alternately as a rod form and as a spherical form, which is due to the fact that its cells are often — as are indeed those of most streptococci — somewhat longer than they are broad. As the earlier descriptions apply in reality to the entire genus Sireplococcus (and largely also to that of Belacoccus), none of the names suggested can claim priority. I believe, however, that it will be in con- formity with even the strictest requirements in this respect to use the name Streptococcus lactis, since the generic term Bacterium cannot be employed in the present instance. Streptococcus lactis is killed at a temperature of 60°—70°. Its optimal temperature is 30°, but it can form just as much acid at 20°. It grows poorly as a rule below 10° or over 40°. In the case of several strains, the maximal temperature is reached already at 38°. No. 16 grows at 42%°, and Nos. 13 and 14 even at 45°. Streptococcus lactis grows extremely fast under favourable conditions. When freshly isolated from milk, it will coagulate sterile milk in less than 24 hours at 30°, forming therein 7—8°/o9 lactic acid. In this state it is also generally capable of dissolving a small quantity of casein. This faculty is, however, in many strains, very soon lost, disappearing with surprising rapidity when they are cultivated on artificial substrates, and they are then but ill able to thrive at all in milk. The lost power can only rarely be restored by regular transference from milk to milk. This feature, and the fact that many strains prefer maltose to lactose, seems to suggest that milk is not the most natural substrate for the present species. Possibly it may be derived from cowdung. It gives milk and cream either a purely acid taste or an unpleasant flavour. | Streplococcus lactis is characterised by its lack of, or extremely slight power to ferment cane sugar. It is likewise incapable of fermenting raffinose, inulin, or (withthe exception of No. 1) starch, but does ferment dextrin and salicin. Like most other lactic acid bacteria, it prefers levulose, glucose and mannose to galactose. Of alcohols, it only ferments man- nite, and not all strains can even ferment this. Its action with regard to pentoses varies. Some strains ferment neither arabinose nor xylose (O-forms); others ferment one of these pentoses, and others again both (4 + X-forms). In Sc. lactis, as in most other lactic acid bacteria, the power of fermenting one or another pentose is generally impaired or alto- gether lost in the course of years; this faculty therefore is not suitable for further subdi- vision of the species; we must as a rule restrict ourselves to noting under each species -O-forms, A-forms, X-forms and À + X-forms. Inits relation to other sugars, Sc. lactis varies only very rarely, but may do so at times, and we may in this respect call attention to the interesting case of No. 12. This strain very soon lost the power of forming acid in milk. Later on, it also lost the power of fermenting galactose, and naturally enough, the fermen- tation of lactose in broth was reduced at the same time. At 20°—30°, Sc. laclis (PI. I—IV) appears as a diplococcus, or in very short chains. In milk, it is almost exclusively a diplococcus often slightly pointed. On agar streaks, it may be elongated and markedly pointed. On AG it now and again forms long chains. At 10°, it always forms long chains, both in broth and on agar streak. At maximal tempe- rature, it either forms long chains of cells having the normal appearance, or short chains of irregularly swollen cells. Streptococcus eremoris (Table XV). I have thus named the lactic acid bacteria first studied by STORCH, which, owing to its aroma formation, has become generally used 55 133 | | | Strepto Salg ale) o\® o|o el! | lol i | | Mil eptococcus = | 2| 1S$/23\2) 42/2/2812) 5/8 sl slalelc|a TNT No.) eremoris 8° 3 | 8 E 22 AE || s SIS : 53|8|82 |3 legiaz Io of isolated from: | 24 (2 28 =5358|8 8841818 85]283 13 ieee, TotalN : = | En | | er an 2 Bega. 2 any S| HE |= SSO 4 | | 6c = 2/<*\sn| DN = STER CN SO D es A ON ea a i W 1,4/2,3| 0 |2,3/2,5| 0 | 0| 0] 0 |1,0| ul 1 Buttermilk1 | & 2188 024361 0 lol ojo 2) 1 Ki 5,0| 1,6 = I ie } | + } Danish dairy” | 5,6/3,6 | 0,9/2,7/6,310,210,5/1,4) 0 13,21 4 | 5,0 8,5) 2,4 ial ‘| sli] Fags TE AS Du | BE 3 rear C 4,514,613,8 | 0,718,714,20,1 | ol1do> 2,513 | 5,2| 04 0,4 4 C [o10|5 44545 1,4 0,211 4702105077 0 091 3 | 6.2 1,5| 0,4 5| Sourmilk 1 | C 0 1542,70 1,1154 0 002101012169 | | i 1 | x fay: an I | | | | 6 RES Cc 0 3,8|2,7 || 0,2/0,7 1807 0508102 36/2165, 31] 0 1 i = ER RE N ed SEER 7 |Home Starter )1| C 2,927 SE 3205 (020502) 0 | 2 |4,7 8 > Il © 3,72, 070,741 0} 0| Of 0 35|1 114 ment LV 1321810 0229 01000 oi | ; = | Starter 4 Cc 6,352 0 |0163| 00|0|10 34 2 ! 3 ze alk I 1 (Ei Ei wirt au LW 2,72.01,610 0212710101010 | | | C 6520 | 0 01591 0 | 0| 010 34) 8 63114 1,8 mt t Esq =o - +—- + + + i w 299,310 01261010 0 0| | |... | 815,4 Buttermilk 2 | Cc 653,8 05,0 6, 0 0 0| Je 3 | ji 5,6 nl w 368,70 052910101010 | | 2| 63) 85 1,2 Buttermilk 1 | W 322,5 | 0 = 2910/0100! |1 [5 ic 14 | Commercial 3 TA NS De Dir | a 14 Starter 7 W 1 2,7 | 0 0,2 2,9 o [ol 0 0 |. 1 59 10,8) 2,8 Buttermilk 2 | W 321322,5' 0 0,2 2,91 0 0/ 0| 0 | BE | 651204 8,0 = hir A| Eee — = Ze i I | Lee nm + = — Commercial | ts | » 16 Starter 8 c| | i mad hea nal 4,7/0,2 10,5 0,2)0,2 0,2| 1 & 17 | Home Starter 1) C 4,3,2,0 | 0302 4,7\0,1 Lo Of 0; 0} 1 17,2! es D —— Ww 342,0) 0 (0,1|2,7| 0 | 0] O| 0 | ae u : 21 ¢ 7,0 54105 0 (6,5! 0 LE 0 0,5| 1 ba 6,7) 0 Swedish w 200,51 0 oılıslolololo jo7/ I | | 19 & || „il, [D | ropy milk 1 Cc 63 47 0,2 0,54,5| 0 | 0| 0! 0 129 1 |6,5,16,5 10,1 English er ee TURN ay ae te ' letras 20 watt’ REG, 5,9150 0 0,515,610,2| 0| 0| 0 0 1 168 le 4,0 Dutch w | bstelooabolo otololol | | | 21 99 45) st |=) | | | ropy whey | ¢ 5438| 0 (0,214,110,2| 0| 0| 0 | o | 2 | 1/101) 7,9 for souring cream in the manufacture of butter. It is therefore found in most commercial starters and in buttermilk. When cultivated at 10°—18°, it often exhibits a tendency to render milk ropy, so that it can be drawn out into threads, and the well known bacterium 1) By commercial starter is meant the commercial culture of lactic acid bacteria, which is used or the ripening of cream. Home starter means such a culture cultivated for some time in the dairy. 134 56 of ,Tätmjôlk” Baclerium lactis longi, and the bacterium of the long whey Streptococcus hollandicus, are only slime-forming varieties of this species. When cultivated at higher temperatures on artificial substrates, they rapidly lose the power of forming slime. Most of the strains of Se. cremoris showed only slight power of resistance, and died off in course of time; a few of the strains, however, were fairly resistant, and one (No. 18) has even kept alive in agar stab for over three years without re-inoculation. This same strain had previously distinguished itself in the butter manufacture, being found practically as a pure culture in a dairy starter which had not been renewed for many years. Streptococcus cremoris is killed by heating to 65°—70°. Its optimal temperature lies below 30°, and the bacterium can, as is well known in practice, be trained to grow at fairly low temperatures, so that it sours rapidly even at 15°. Very powerful strains (especially those forming slime) may develop at 3°; others less strong will not grow at under 10°. It does not generally grow at blood heat. In most strains, the maximal temperature lies at lores Streptococcus cremoris rarely forms much over 7 °/,, lactic acid in milk. It can develop such a quantity of carbonic acid that fine stripes appear in the curd (illustration of fer- mentation test g. 2). Most strains have a certain power of dissolving casein; No. 15 has even formed over 20% SN. It does not thrive so well on artificial substrates as Sc. lactis, and thus proves itself a more pronounced milk bacterium. In accordance with this, we also find that among disaccharides, it attacks chiefly lactose, and prefers this sugar at the concentration at which it is found in milk, or even higher. Saccharose is practically speaking not fermented at all, and maltose, and thus also dextrin, only exceptionally to any considerable extent. The best starters seem to have the least power of fermenting these sugars. Some strains ferment salicin, others not. The three monosaccharides, levu- lose, glucose and mannose are on the whole fermented equally well. Alcohols and pentoses are as arule not attacked at all (with good nitrogen sources, there may be a slight sugges- tion of fermentation of arabinose). In morphological respects, Sc. cremoris (Pl. IV—IX) differs from Se. lactis by forming chains, often of considerable length, at optimal temperature, both in milk and broth. The more powerful the bacteria, the thicker — and generally also longer —- are the chains. Weakened strains can, in broth, form chains so twisted and clustered together that they may be mistaken for staphylococci (PI. VI, No. 18). On agar streak, the cells as a rule assume a very abnormal appearance, becoming swollen, pointed, and often rod-shaped. The rods can, however, in many cases be resolved into discs on further observation, and then resemble woodlice or other articulata. On agar streak (at the optimal temperature), we also often find division in the longitudinal direction of the chains. (PI. VIII, No. 20). At maximal temperature, chain formation ceases (PI. IX, No. 2). In gelatin, the chains are often short, and the cells (especially on AG) irregular. As regards slime formation in milk, this is due to a slimy transformation of the highly swollen capsules, which may be found in all lactic acid bacteria in a young state, but is particularly frequent in this species. When the milk becomes so sour as to curdle, the slime formation ceases again. No. 19 (Pl. VIT) shows this development distinctly. If a stained preparation of the bac- teria in the capsule stage be exposed to pressure, the streptococcus chains will burst out from the capsules, appearing then as white worms. 57 135 Streptococeus mastitidis (Table XVI). This bacterium, which is also called Se. agalactiae, produces mastitis in animals and man, but does not appear to be otherwise pathogenic?). Table XVI. | ie lolel | A lo | | Milk | © eg 2 ia} | 21 © © | o =| MM MESA == Streptooocens SE VE ls) 2/2) 2/2/23) 3| 3818 3 8 Bla lssle leet eel % of No. mastitidis (este Salat al ls1S$5|8) 0181221815 |218418 108184] : lot fini | 22) 2 he) «ss 5] a 2 BØDE = | 2 S £ = £ | © = c: 22 23 Total N solate om: | Sale || Bl Sl Als | lals | | = = IA A 5S Eu FRA | 2 Ele ae | AE = 2 8|2|% =| = (92 ES Total N | isolated from: | SENT || £| 2/8/51 8/28 = CARRE J|Sl|=|Alni EEE M | ke © <ım Aal] | | É 051805052025 (50438820 49383829 3 23342082) 2 | 4,7| | — | w 10,2] 0] o| 00 0 (25282714 16251416 16181118 | me En (°° 0225! 0 0! 0/0,9 |7,06,86,8 5,0 | 5,9 6,3 5,9/5,4 |4,74,3/1,62,9| ? | 5] 1,0/+0,7 : à .[w] 0,5) 0/01 0 ofa) 232083011 20202018 201,6 016) , baad == PB eet Se) | € 10723] 0! 0! 081545456441! 615664541433, 0 34] ae 6] 4 | wio 0/0 0 01,8] 2.5[2,3)2,92,3 | 1,8)2,412,41,8] 21, | [2520703 7 » 72 | » josızlo| o| of20! 2712, 2,72,8| 2,52,5 2.420 |: 23 | | 2 |5,9,0,1+0,3 8 | Chinesehen’segg| C (2,5/2,52,5 0,7/3,2 | 4,5/4,7/4,514,0 | 11501529 [27 72,7l1,114,11 2 47147 0,6 to 60°. Its optimal temperature is 30°, and it thrives well at indoor temperature, may ‘indeed grow at 5°. Nos. 2 and 3 do not grow at 40°, whereas the others do at 45°. It is difficult to keep alive. Sc. inulinaceus differs from Sc. bovis in not attacking casein and in fermenting inulin and raffinose, even with W as source of nitrogen. It often ferments starch and mannite (No. 3 even sorbite). All strains with the exception of No. 1 will, with suitable nitrogenous food, ferment xylose. One strain (No. 8) which was found abundantly in some dried white of egg (hens egg) from China, was also found to ferment arabinose and glycerin?). Sc. inulinaceus forms for the most part short chains on all substrates. It has no cap- sule in milk. On agar streak, the cells are often markedly elongated. In stab cultures, there is no indication of surface growth. ') The East Asiatic Company, of Copenhagen, made some experiments in China with a view to the production of dried eggs. When concentrated in a vacuum at 30°—40° the white of egg turned to a slime, which held the water very strongly. In this slime, we found the above-mentioned inulin- fermenting coccus, together with a coli form. These bacteria do not appear, however, to be responsible, either separately or together, for the slime formation. 61 139 In their attitude toward the various sugars, Sc. bovis and Sc. inulinaceus are so clo- sely related that they might be regarded as A- and X-forms of the sam especies. As, however, they also differ in so many other respects, I have considered it most correct to establish two distinct species, though I must admit that in the case of some intermediate forms (0-forms) it is difficult to say where they should be placed. Streptococcus faecium (Table XX) is the most frequently occurring streptococcus in the human feces. It isalso found in the feces of other mammals, and in this respect, strains Table XX. | | 1} 1 = I Fr Le || © | 2 Milk | Streptococcus [SS]: 22 g 2 = 2 2 2 E $ 2 Ne =!&|s|al 0 | ole of No. fæcium polo Se SE el els El SIS SIS] SEN SI SE, Siege] _° | isolated from: 32 AE = ‘= = @| 5 = 8 = Ib E 2 3 c 25/82 Total n; RE lak el REE | a nal SHES pester 1 Fæces 2 Cc 110836348484 63635042! DEREN 0 1,602 4,1 5150] 16 03 = Se | EL, = ! 1 2 } | 2 » 2 |» |ososk13453483| (63615048) 05504305) ¢ 016024113 159 11 03 3 SEE (0,2/0,715,810,5 542,9] 61665481 4,75,04,12,30,511,4.0,54,1| 750) 58 03 4 » à |» (0,7 0,7 4,312,334 3,2 6,8 6,1 5,4 4,1 6,215,04,3.3,6, 0140236 5152] 4,5) 0,3 5 sos fe mnt el 4,54,74,106 0 rn [52] 21] 0 6 » 5 | > 0,7. 0 |4,5/3,4/0,2/4,1 7,1/7,2/6,8 4,5 5.4168 5,2 4,7 0 2,510,215,4 4 | 5,4) Schwedish |, 10,210,2\6,611,8| 0/3,3| 1874 685,0 45745409 0125 015,9) 6 145 Yoghurt || | | ua au | al i AE ol: ie che | ah Milk left to stand | W 10,8] 0 /2,7/0,5| 0 bol oar 123209 0 0| 1,6 | 8 "9 | wh | || | 3 . donne aol Cc [os 0 31180288] 1,068 6,35, [58655641 024 ( 11052 316% 4,1) 06 À er =» es TT —- Tr So d t t es | | | ure ae | » |0,7/0,5/4,0)2,1/0,5 52,7 5450413814552 413001140242 3,6| | E | ee sp | 124 Eu ee: Soured potatoes ate 1 | 1 1 T i | — | Al Fæces 1 0,7/0,7/3,7/2,00,5.2,7 | 5,4 4,34,1 2,3 |5,65,04,12,80,1/1,30,24,0 | 3,6 1 a Seas tp = 1,111,613,211,110,5 2,5 | 5,6 56 54/32 |2,94.5 362, 007 7/2, 5 0, 7 3,2) a 152] 3,6|-:0,3 mus : ha sah Feces from bottle fed child I Feces 2 | | u 2 0,91,613,6.020,234 |6,56,86,54,1 |4,76,1/5,24, 30,2 3,2 ne 4, 7 9 | 4,31 0,710,9/4,10,710,5/4,2 | 6,85.915,9 4,3 15.215,04, 13.40511,4 05 3, al als 45 04 i Ps] 7l eo 2 | | | Dairy cheese 2 P| W |0,S0,1/2,7, 0! 0 MES | 0 36,0) 0! 0 | 6 14,7] 4,4|+-0,2 —— 0,50,2/5,2| 0| 0 3,4 7,07,07,04,5 | 0,96 8/54 0,50,22,0 05,9, Feces 2 a 67 48 0547 43030514 0 34 2 |6,3| 42) 03 0,5 i an 0750 00202 68596141. 0,516,014,710,710,51,4 0,545) 2 54] à 1,4/0,2/0,5 0] 0] 0 150505225 /014,72,70,5, 018 21660 sel 16 0 0 (0,2 | 9,0,9,28,8)5,9 | 0,5|8,6/6,5/3,210,2/3,4) 06,5) | | | ” Calf fæces 6 —+ —— + ; = —+ of me, TEE eee san ol | | fs [ox OT aie Düggeli eh Ua 0! > | = 18; fe „ia | al Ran | “Faces from arc- iis 1. : | | | 4 19 | tic fox. 1,634 4 37 2,3 3,6 3,8 6,1/5,6,4,7 4, 3! | 5:9) 4,74,71 0; 42 230542 2 | 4,7| 5,6) 0,2 | North Greenland | Ed etd ae eae! | cl eee fe oe) SE Se ES ES i ende N | Feces from | | D LS] a seal. N; | TREN 6,5/5,2 3,8 4,5 | 5,2/5,4/4,7 07 0,5 1,8/0,9 4,7 2,7 -:.0,4|-2-1,6 | | ll | = North Greenland H = | 140 62 Nos. 19 and 20 are particularly interesting, having been isolated from feces which were extracted with due precautions as to sterility from the rectum of a blue fox and a seal (Phoca foetida) in the extreme north of Greenland, at a spot where infection from without would be most unlikely to take place”). Sc. fæcium must therefore be one of the commonest intestinal bacteria among animals. It has a high power of resistance, and not a single one of the strains has died out under our hands during the years we have had them. It is also fairly omnivorous, and can grow at widely different temperatures, so that it should be able to thrive practically everywhere throughout the world. It can stand heating to 70°—75°. All strains develop at 10°, and some even at 5°. It exhibits lively growth right up to 50°. Exceptions are the two strains (17 and 18) with slight fermentation of arabinose, which do not thrive over 47 4°. Owing to its good growth at 40°—50°, it may be found in milk which has been kept at high temperatures, as with Nos. 7,8 and 18. For the same reason, it is found in vegetable matter stored warmly in pits, as Nos. 9 and 10, which were found in boiled potatoes which had soured spontane- ously. Streptococcus fecium grows rapidly in broth, but comparatively slowly in milk and never attacks casein to any considerable degree. It always ferments arabinose, but only exceptionally (Nos. 11,12 and 19) any great quantity of xylose. It generally ferments mannite and saccharose, but in respect of this faculty, fluctuations may be observed even in one and the same strain. No.1, for instance, on the first investigation, fermented sac- charose, and No. 15 mannite, but by the very next investigation, they had permanently lost this power. In their relation to the sugars, the non-saccharose-fermenting strains resemble Streptococcus lactis. Nos. 1—4 ferment sorbite, and are thus nearer the species next following, Sc. glycerinaceus, Sc. fæcium often ferments rhamnose and raffinose, but in this respect, we may also find differences within one and the same strain. The present species is thus characterised by a comparatively wide range of variation, and it is there- fore not remarkable that strains of the same origin (Nos. 2, 3, 5, 13 and 15) should exhi- bit such differences in respect of fermentation power as shown in the table. Streplococcus fecium is a pronounced diplococcus, (Pl. XV—-XVI) which does not as a rule stretch before division, and may therefore — especially on solid substrates — _ present a micrococcus-like impression. Even in broth, it only forms very short chains. Some few strains (6 and 12) can form elongated cells on AG. Streptococcus glycerinaceus (Table XXI). I have named it from its remarkably powerful fermentation of glycerin. The few strains which we have succeeded in isolating were all with one exception derived from cheese. As it grows, in contrast to most true lactic acid bacteria, equally well on AG and SG, its presence in a substance is most easily discerned by spreading a little on AG. In stab cultures, it exhibits a somewhat stronger tendency to surface growth than the other lactic acid bacteria. Like Streptococcus fæcium, it can often stand heating to 70°—75°. It grows between 10°—45°, and two of them (6 and 7) even developed at 50°. ') These strains arise from an expedition of Knup Rasmussen. The samples of faces were taken by the Swedish botanist, the late Dr. TuorıLD Wurrr, and sent to Prof. BARTHEL in Stockholm for a complete bacteriological investigation. The isolated lactic acid bacteria were further sent to me to be exactly identified. 63 141 Table XXI | | | | | | “| | : | | © | © | | Milk Streptocooons |) 55 8 521213 32% 2 2 selle el on of i saleıciEelElS|=|j<2| 5/5 Sal 82 23|5|35 2/8 Følg] _" No... glycerinaceus 581212355 S21 8| si) 2/4\s SE 4 513 [033% Total N isolated from jodi) oi ali s|i ald sj o|s Simi Alan | ses if al ‘al Beds) aia | | |ÉSKSIsN) DN ie |. CR —- + | i Po je | 4 | Erst fe T (ee | RE sa Date al w [1,5 = 04! |2,311,7 |3,7/4,314,1/2,6 0,53,4 2,9l0,1 | 0 | | 5 [54/53] 03 | ? | | | | « ’ Ose ” 1 month | C 25 2,7 0/48 4,3/4,7 | 6,5 8,3/6,5/5,4 |0,516,5 5,910,2 | 0 2,7 0,2 6,9) : | = | FOR 2 0 EN A | : RUE eur W 11,6 x 04) |2,5/2,3 3,814,3 4,1/2,5 |0,513,4/2,910,2| O| | 9 Sol lan ? | Vol Soe (rs ’ 4 months | C |2,5/2,7| 0 13,314,5/4,7 | 6,5[7,7|7,015,2 10,7 6,115,6 0,2 |0,2 34 015,9 | | Dairy ch | 200,1 ISA A CN NN EU TE | airy cheese | wy 1529197) 19,30 5|3,714,314,52,5 12,013,412,910,4 10,51 | i 3 9R I” 10,5] 0 ails et IR 4147103| 0 > | | ’ | s 1 month C 12,511,1| 0 |3,014,3|5,0 | 7,717,9 1,236 17,0 6,5/5,4 0,7 0,5 250,215, zi | Dairy cheese | w h 6801! asbalsraıısesasgeroslo | | | |... | 4 | 9R 0,8 - | | | 4 15,0134} 0 3 months © 12,512,7| 012,04,5/5,0 | 7,7/7,9|7,7|5,4 |7,016,515,9 0,5 0,512,7| 0 611 | | al "|" ’ DEAN] et pa et a a | | | Fæces from | | | | | 5| bottle fed c [2,3 al 3,8/3,8 || 6,316,5/7,814,1 15,26,1/5,20,9 10,212,90,515,6 | 1114,7| | child I | DR | | | Be I Joaloa | CE SE NØ oO RARE bat | 6 vei es WIL | 12318313,6423,82,213236260,4| 0 LE lear | 610, 8 3 months | C 2,7 0| 0 1,13,44,5 |7,9/7,4.7,7 3,8 16,3/7,0,5,40,7 0,5 2,1] 0150! | | | Surface of the | | lite] | 7 | Camembert c 240,714] |3,213,4 || 5,5/5,415,213,4 |3,2/4,7/3,7 0,8 10,62,511,014,6| 6 | 4,3 cheese 4 | | | | | hl) | | | | Sc. glycerinaceus always takes several days to coagulate milk at 30°, and it has like- wise no pronounced power of attacking casein. It grows better in sugar broth, and grows not only, as mentioned, in sugar-free gelatin, but also in sugar-free broth, so that it is capable of developing in cheese after the lactose has been fermented. As a powerful fermenter of glycerin, its growth in cheese is furthered by the fat-splitting process which there takes place. It ferments not only glycerin, but also other alcohols, such as mannite and sorbite, and even shows some indication of fermenting dulcite. Still more remarkable is the fact that some strains (3 and 4) can form up to 2°/,, lactic acid from inosite. It always ferments rhamnose, and often xylose. The bacteria coming under this head have shown a remarkably constant power of fermentation with all sugars, except the pentoses, a perceptible decline being here observed during the years we have had them under cultivation); we are therefore also disinclined to attach great importance to the difference in power of fermenting xylose shown by the strains investigated. In contrast to the remaining strains, Nos. 1 and 2 have an extremely slight fermentation of saccharose. As regards the polysaccharides, dextrin is the only one which Sc. glycerinaceus ferments to any considerable degree. 1) The quantities of acid noted in the tables were in W derived from the freshly isolated cultures, wherefore there is sometimes a higher fermentatio of pentose in W than in €. 142 64 In this group (Pl. XVII—XVIII), we encounter both pronounced diplococci (1, 3, 4 and 7) and pronounced streptococci (2, 5 and 6); even strains which appear identical in biological respects (1 and 2, for instance, which were, moreover, isolated from the same cheese) may differ widely as regards their morphological features, the difference here being maintained throughout a period of years. No. 6 (Pl. XVIII) forms particularly long, tang- led chains and thus makes a very typical flaked precipitate in broth. The diplococcus-like forms in streak cultures, and especially at maximal temperature, exhibit markedly pointed cells (No. 4, PI. XVII). Streptococcus liquefaciens (Table XXII) liquefies gelatine and peptonises milk strongly, especially when the acid formed therein is neutralised with chalk. Other- wise, in cultural and morphological respects, as well as in its relation to temperature and Table XXII. 7 a BR a Wee ae ie Milk | Streptococeus esi 2 2 8 21216 2 2/3 SEE SlE/S| Si ls | 9%. of No. liquefaciens |e2 | 8/2383 8|33353% 532 2/8|5 8) 8/2 |28]52 Total N | isolated from SE glx} g | Sse COPIE 8/S|S\4 SA |A ER EN pen | | “= "| Wie: = 1FS | °| 0] 0| 0| 0] 0 4,115,04,52,8 16,241) 0| 0 | 0 455232 0,2 - ton at Ir =r | + SC | ol t |. Epizootic | | | | | | Ip i eal | | | | | ortho ect FE KU 010210 41 0145544738 544545 0 | 0 505634 (1,6 | C. 0. Jensen | | fi | = ‚are | | | | bag | Petechial Er | | | | 1 1 a i | | | i 4 | of the horse » 10! 0! 0! 01/43! 0 4,3 5,0 4,5 3,6 5,24,74,3 0 | 0 |5,0 5,6 3,6 C. O. Jensen BE, | + + 1 + + - i a} Navel and joint I | | | 1 | | | | 5 | evil of foals | > | 0/0 0) 0/41) 0 435,044,534) 5,45,04,3 0 | 0 44,75,63,8! C. O. Jensen | | | | Fe Sc. glycerinaceus BD |. 1 | Rå pot NAN er 23 0 al 3,8 | 6,3/6,5/6,8/4,1 | 5,2/6,1/5,2/0,9 | 0,2/2,9 0,515.6 | ae | RUE : | eo: {7 | ED 1H Facial erysipelas | | | | |. ote Oluf Thomsen | > På 0 0| 0! 013,8 384,74,12,7 | 5,014,332 0 | 0 270,734 ——- tt 1 | 11-13 — = N | = Sc. mouse | Le) | wa Léa 7 | vil Jensen. | > 101005 0| 02,9) 6,615,6(6,214,3] 5,414,310,21 0 | 0 27 0152 | z I | 1 Sek | Å LI eM bed ui =| Sour cabbage | 174 | 17541 La sé sl Lay 1 » |0,2/0,20,2| 0| O 09! 5,6 5,415,334 5,213,4 0,6) 010 23 05,3 2 Page S| Le 1 — Tr =! . i Sc. rabbit T ae 1 | | T | Vil Jenan | > 090205 0| 00,2) 4,54,514,33,2/4,715,2| 0105! 0 br 0.29 € | ERA PRE i | ~ | Sc. pus Al | + RE | 1 | Vilh. Jensen |» 1090205] 0| 0 ia haa 5050 0/0,5/ 0 27029 lol » a a ie ue Me Sc. fecium b | | 2 d Le | No deel ee saone 6568654147161 5245 02820241 9 14,3 J | | = 1 ES We Lt N = LER Se. new I I | li |_| | TD EN EU" tt 10 | “pith: dena” Me 50,254 0 bape LL 7016515638) 0 2,3 05,9 5 4,7 i; + || i | + | | + | N — 4 En || —_— Se. Fredericia | | | ga rere "sd | | 11 | Yun. Jensen | ? 050256 o naja [7,274 048] T1606) 0 25 0/59 6 43, Sc. fecium No. 6) » 0,7 0 (4,5/3,4/0,2/4,| | 7,7/7,2 6,8 4,5 | 5,4)6,8)5,2)4,7 | 0 25 0154| 4 15,4! | L a i 3 67 145 son with our saprophytic streptococci, we have investigated, by es same methods, some few pathogenic streptococci, kindly furnished by Professor, Dr. C. O. JENSEN, Dr. WiL- HELM JENSEN and Dr. OLUF THOMSEN. All of them form dextro- De acid without any considerable quantity of by-products. Nos. 1—9, which appear in all nutritive substrates as shorter or longer chains, will not stand heating beyond 60°, and thrive very poorly below 15° and over 40°. Nos. 10 and 11, on the other hand, which in all nutritive substrates appear chiefly as diplococci, are not killed until 75°, and grow well at temperatures from 12° to 50°. They are thus not affected even by the highest fever temperatures. Nos. 8—11 were not difficult to keep alive with the nutritive substrates employed, whereas most of the other pathogenic streptococci died off in course of time. As regards the chain-forming strains, it would be natural to consider them related to Sc. mastilidis, which forms the proper connecting link between the streptococci of milk and the true pathogenic streptococci; they resemble this form also in growing particularly badly in yeast extract. Some of them also ferment starch (without attacking raffinose and inulin), which was, as we have seen, a particular characteristic in the freshly isolated strains of Sc. mastitidis. They form no colouring matter from it, however, and differ per- ceptibly from Sc. maslilidis by growing poorly in milk; several of them, indeed, do not ferment lactose at ali. Like the other long-chained forms, Sc. mastilidis, Sc. cremoris, Sc. thermophilus and Sc. glycerinaceus No.6, they do not ferment pentoses. The strains which produce epizootic pneumonia, as well as petechial fever of the horse and navel and joint evil of foals are distinguished by fermenting sorbite, without being able to attack mannite, which is otherwise far easier to ferment. No. 8, and especially No. 9, are red in the stab in casein peptone agar, a feature which we have not observed in other lactic-acid bacteria, but which is said to be characteristic of Sc. lanceolatus and Sc. mucosus and which is very common in the propionic acid bacteria. There can hardly be any doubt that the present group contains several new species. In the table, we have shown the fermen- tation figures for those of our saprophytic streptococci which most nearly resemble them. The one isolated from sour cabbage is a long-chained streptococcus which does not grow at temperatures over 40°. As regards Nos. 11 and 12 these should, from their appearance, their relation to temperature, and their fermentation of sugars, simply be regarded as pathogenic strains of Sc. fecium. It follows then, that the name Streptococcus pyogenes — like most of the bacteria names hitherto employed —- is a collective term, embracing several species. WınsLow ') sets up no less than six species, three of which do not resemble any of our strains. With regard to Sc .lanceolatus, of which we have also investigated one or two strains, and Sc. mucosus, these form no considerable quantity of lactic acid at all. Judging from their shape, we may just as well reckon them among the micrococci, and WınsLow also ” places them in the genus Diplococcus, which comprises the most pathogenic micrococci, 1) Systematic Relationships of the Coccaceae. New York 1908. 19* Genus Betacoccus (Abbr. Bo). As mentioned in the introduction to the streptococci, the betacocci are found in green vegetable matter and juicy roots. They are as widely distributed in beets as the saccharo- myces on sweet and juicy fruits. They are introduced with vegetable food into the intes- tinal canal of animals, and pass thence into the milk. In the retting process we always encounter arabinose-fermenting betacocci, which might be connected with the fact that pectin substances always contain an arabinose group’). As these bacteriacan make vegetable matter tender, e. g. sour cabbage, it is possible that they play some part in the retting process itself; this point, however, we have not yet been able to elucidate thoroughly. As the betacocci are far more variable in all respects than the streptococci, it is very dif- ficult to divide them up into clearly defined species, and I therefore prefer to treat the genus Belacoccus under one head, and merely note in conclusion what features might seem to justify our uniting certain strains into independent species, In Tables XXV a and band XXVI, they are arranged principally according to their relation to the pentoses. The betacocci can as a rule stand heating to 60°, but rarely to 65°. In a slimy state, however, they can stand higher temperatures, as the slime protects them, and it has been observed at sugar factories that thin syrup which had been heated to 80°—85°, and could not possibly have become infected afterwards, has grown slimy (No. 11 has formed -zoogloea masses under such conditions). The optimal temperature lies at about 30° or under, a single strain (No. 14) was even found to grow best at indoor temperature, and this temperature is, asin the case of Sc. cremoris, the most favourable one for slime forma- tion. The maximal temperature is 35°—37° (rarely 40°) and the minimal 5°—7°. Some few strains (Nos. 1, 45, 46 and 47) grow, however, at 45°, but on the other hand thrive but poorly below 15°. The betacocei always form levo-lactic acid, more rarely also equivalent quantities of dextro-lactic acid, so that we find inactive lactic acid (Nos. 43—-47). As certain strains (Nos. 6, 41 and 42) in a freshly isolated state formed inactive lactic acid, and later only lævo-lactic acid, this is evidently a variable quality, which cannot be used by itself as a species character. The betacocci also as a rule develop gas, (carbonic acid with more or less hydrogen). The gas development is strongest in lævulose solutions, and next in cane ') FELIX EHRLICH: Chem. Zeitung 1917, 41, p 197. 69 147 Table XXVa | B „= 1) pe] | T = " Betacoceus |2 =. | © | | | | | = == 5S al a] ol 22 o | 2 o|| 3 | | | Mik _ vo, Aranda + PP 2 2 2/2/22 5 E 8/8) 2)2 2/4 elas | Æ peus AE SEHE 2 &' 8 8/8 8 ET 8 8 595. of lees isolat Bole alloy) ale\s €) 2) #85 02 53 EE pete from Bar| HEISEINEIEE HEIEE 2/4/23 Fe ER Total N. ÆG så Fa AE nf oo) | eS ERE RGR Se ee Be | | Emmental |; w | 0 (0,7/3,2) 0 | 0 0,512,912,5 2,5|2,5/2,3|1,4|2,3 al olo! af 01 | cheese G | 0 0,9199] 0 | 0 0.965,63 5.04,711.411,4 6,1 ob2l02! olos| 6 | 54l Y | 0 10,9/6,8| 0 bal 210,2! 0 0,51 6 | 5,4/<+1,2/+0,4| 0 79/6, 0 0,915,6 5,614,7 a = = + | a 32192179222 4,112,312,916,1| Ol! 0 0,2] 0 O | | 2 Bovinefæces1 | 1 C |0,5! 0 (4,310,7 | ARS DEA SA HT. | it a es 102058,62,92,711,61,62,72,92,50,20,50,50,2' | 0,7 0 » — ul 2 | IK 020,7 3,6.0,5| 0 0227 1,8|2,3|0,9/2,3/1,8)1,1/1,1) 0 0,505 (OR ial i 0 Dairy cheese Ww 0) 0 4,7 0 | 0 | nm 4 co aa | 1,8 0) 0NL12,0/L8114, 11111,11,11 0) 01 0 | C |0,2 | | 3 months ,2| 017,7! 0! 0! 0 16,015,2[5,413,211,612,014,311,11 0 | 0} 0 18| 6 | 5,5|+1,6| 0 m | LILY 0|05,2 0| 0! O|| 16,8] 14,115,616,5 al 0! 0! 0! ae Dairy cheese = | | ER ARE 18 nues te fig Me TR N 0/14/38 0| ol 0/2,7/2,712,7/2,7 232723 ]4 0 010! | | | | 4 months c | 010.9179, o! 0! 065 0,2 | 2 | 7,2 28 | ta at | 200 | 16, eas 5,211,614,114,311,110,200,2| 0 4,5 | 9 | 5,4) © 0,6 | m \ Dairy cheese | ; W | 010,9 oy 0! 0 0,714,312,711,8 2354 1 eh cool] 0 0 Be logie” 9R 0 A We G | 017,2 ‘ eee 26.8) | 911616 i lg 015,612,0/0,7| 01 0| 0] 0/4,1] 4 | 54.08.08 à | | 2,8--0,6, MI — L 0 1,1145 HR 1,6| [61 [8,6150 Sa 0041 2,0 ) w 0094 0| 0 He fo Dairy cheese |” if 0 | 2} 0| 018,213,911,812,08,711,212,01 0} 0! 0! © 0 Er 9P 1lc | ‘ol | enc 0| 07,015,4/7,0 3,216,818,115,00,21 0| 0) 0) 0 | 3 10,4 | Y | 8 | 3,4 07| 0 | MI [| 0 | 0! (6,1) | 15,2 071010) 0. 0 | i tr + ++ Fi BER | | De | BIER N Mn cr oo oo. || C | = | I2|68) | | I _ BER HER SAS SE GEG ale al le al | MEE SELE Na 017,9 0} 0 0! 0! 0 | | | wl 0 7 1 + —+—- = | | 9 » 1 0 0,9 4,3/2,9/2,012,3 3,6 2,0 2 = 0! 0| 02,51 | #0, FES RE 10 lis | i : ee Woes alot 0 45) (27 27 | 16 | M À. TI pen > ? 6 2,912,3 0 | j | | 10 | Kefir" 2 i C 3 el >| 2 —t 9 | b) | O12 | 0 4,5 | | | St 5 015,2/1,6'2,0! 0 |4,311,6| 0 0,21 0! 0! 0/0 7 0! 14,114,13,415,04,715,0| 9} 011,61 0 0 | 2,5 | ? | MI Slimy thi | N | | he LL | | Dole oa | | | + | heated". Y 0 9027159 1,116,14,111,8/0,6! 0) 010411 | 2,3 ji i | gra 3019; 3810,5/5,0/2,9/1,8,0,7| 0 | 0! 0 4,1| KJ + | Il Sli | See | | Slimy raw W Ar | Ss pan + ” 12 „Juice from 5 à 015,9 2,0 6,8 4,5/2,5 5,4 1,6/2,7| 0 0,5) 0 2,0 | rer ly à 0 17,0) | 1681 4565 0| 0| 05,0 | | Mm 0,9) |72 6,3 5,2] 0 0,2) 0 5,0 æra Mi 148 70 Table XXV b | ag [53] | [El 4 | | | | | | Pr} | Mik |, | Betacoceus 2° | 2 | © o|| a | l= Siew FE 2 la-andA + X| 22/98] 5] | S| 8/2) 2) 3] 318] 2128| 8 ste LE | om off Zac No.|| : l>gios! 9| S| 2) ¢ 21813 es al» = a |. ose Ess forms \5= 325 2|s SE 5 s| à sa: = Sie Ki = 225 55 TotalN.| 223 = 15 de al os « a | S|) || | Ce! ant [isolated El "| 5 2 =z ie 3/8) 2) 5] 3) | [2] SHE D Er | ©| | | | Et 13 | Rotten | W || 0! 0,8) 2,3) 0} 0 0 2,3 = | d swede II d | “| G 10,2 8,8 10,4 0 00 dE | 14 Rotten | D W | 2,5| 4,7; 0} 0 0 | swede IV | |G|0 El 8,3) 0 0,2) 0,5 2,9 is | Rotten || | wi] 0! 0,5! 6,3) ol 0 0 2,0 ° swede |" |c | 0| 721108 0| 0| 04,1 ig | Rotten | W | 0) 0,5} 2,9) 0} 0 0 2,0) : swede I | l'clo 6,1)10,4) 010 0 4,1! = Soured | | | | | | 17 | potatoes 11 | 1 C | 0 7,5 5,5 0,1 Ée 0,1 3,3 | Soured | | 8 | | | 18 | notatogs 211 | 1|c|0 1103 10,1 0,3 0,1 0,3/6,3/4,7 2,03,415,313,3 3, | | 0,13,2 | ee a 52 ARTS | | | E 19 | cabbage aft qe c 01 34 4502102) 0 5,4 524167138507 0! 011,6 20] >.> 21 bi 6,9 79011607 5948453550 868845 10,5 02,8) a | » '» - 211 |e loll 52 9,0 0,1/0,510,2/5,9)2.9 3,6|2.7 5,6 4,3 4,35,0 20,5 0 2,5 = ae ne + + ——- 2| >» > 1/1 | C 0,1) 68 79020,5/0,56,1/5,04,3/3,65,2/5,0/4,5/5,0/0,7 0,1) 0/34| 23 | > > 2/8 | € (0.1) 7,1) 7,7(0,1)0,2/0,3 6,1]5,4/5,0/3,6)5,4)5,4/4.5 5,7/0,2 0,6 02,9! 24 |.» » 1|1|cl01 3,7, 5,4 0,2/0,3)1,1/5,9/4,5 4,1 2,9]4,7/3,6 1,6 4,5 10,7 0,1 0 3,6) 25 NET lc. (0,1! 4,7] 5,90,1 021, 5/6,0/4,14,1/2,5/5,4 4,510,510,2/0,7 0! 0 3,4] : 26 | » + COUT | (2: 2 ; 1) 50 8,8 0,10,20,315,95214,3 3,8 5,4|5,2 3,614,7/0,2 0,3 0 /3,8 | 27 | eo = all] c 0,1 83 8,10,110,20,26, ‚85, 5,04, 35202060302 | | | | | | sugar solutions. Under these conditions, we find a considerably greater quantity of acetic . acid formed than in dextrose solutions. In levulose solutions, the strongest gas-developers (A + X-forms) also produce some mannite. In most strains, however, the gas development is so slight that it can only be observed by sowing out strongly in tall sugar agar tubes, and a few strains (Nos. 1, 2, 3, 28, 36, 37 and 43—47) do not seem to develop gas at all. Several of these last grow chiefly in the upper part of the agar tube, though they are not otherwise obligatorily aerobic. In stab cultures, for instance, they do not show real surface growth, any more than the other betacocci. The gas-developing strains all form more (M) or less(m) slime (mucosus) from cane sugar. These which do not develop gas on the other hand form (except No. 28) no (0) slime; some of them indeed (as most of those forming inactive lactic acid) hardly attack cane sugar at all. In accordance with the investigations of ZETTNow!), we have found that the slimeis only formed from cane sugar (though some strains show an indication of slime formation with raffinose), and ') Zeitschrift für Hygiene 1907, p. 154. x 149 Tabel XXVI. | | | | | saccharose gelatin Bis is ie | B z | en 2 | | | | Milk Betacoceus Bal. al [ele olslolelelstol, | | 1 u 2) =\0|8 190) 2| aw “i | € | ce 2 wo} S7 and O- PLATE 8) 2) 2) RES) S 3/3) 8) 8) 8) SE SiS lewis | co of | forms BSE S| >| ml 2 = B/S) RIS SIS 218] 3) 8 815 04/52! Total N. | É isolated from |£ 219 73 | * E 3 a Ss) aa = 8 AC AI ain EH i g s= | | aid Ihe? Irre] IE | : dd = rt L | 1 ee lie t- te | | É | i sn |pn|< 28 RTS oii tele ER 0| 0! 0/5,9/3,6|3,6/2,1/5,6/3,913,3/4,2' 0103 01341 | 2,0 IL | ‘RER | 22 Ju x si 29 | Bovinetæces4| 1 | C 02152] 0 0,2) 0) 0 5,93,43,62,3 sah 32 3805 230,525] 2 |83 7,9| 1,2 | SO eee C 026,3 0 0,2) 0| 0/5,92,9 3,6 2,8 5,6 2,02,7 4,110,2/1,60,51,8) |1,1 31] » _»5}1}C0,54,7 0 0,5) 0| 015,613,43,412,515,013,413,214,510,511,810,5,3,2 9 |7,4) 0,7, 0 32 | Calf fæces 2 | 1 | C | 0 4,50,50,502 0 4,3/2,92,5/1,4 3,4/2,3)2,5 34 0,5/1,8 0,5 2,0" 2 | 8,6 Milk É 3,2 0,9 w| 00,2 ( : > ne Le 0,2! 000/036 2,9 L838 30° 20! 010!0!0 DA eos c| 0/8,00,9] 0| 0] 0/5,915,015,014,715,614,5)0,515,2) 0105| 0| 01 105 | 06 No. ¥ | 01801 0! 0| 0} ok614,7| 0! 0) 014,707 01 01201 0| 0 Dairy cheese W || 012,01 0| 0! 0! 0 8,811,811,4/1,4.3,2.0,9/0,711,81 0| 0! 0| 0 23 | 34 9R 1 | © 0/7,410,9| 0| 0! 0 6,115,914,315,27,215,615,215,4| 0 10,9] 0 10,5! 7 | 7,7 NDR 3 months | | ¥ | 02,3 0| 0| 0| 0.5,96,3[6,811,116,815,6,0,57,4 0 0,1) 0| 0 Bzw spa 0| 0| 013,013,513,012,011,712,4 1,6 0.0 0) of ol i 35 É 5P É 1 | clo mg 0| 010,2/5,9/5,614,313,614,74,512,5| 0! 0 0,2! 0| o| * = 0,6| 1,4 a months > ;3| | | | ¥ || 0} 0 - 0 16,815,115,613,6| 0 [6,9 0| 010) 0/0 0 | Saito 2,811,7 W| 0} 0] 0| 0] 0} 0302,11] >2,38312,31 01 0|0|0| 0 | 36 | Sourmilk 3 | 1 A HGSILON | ee 4 15,0! 09| 0 . C| 0] 0| 0] 0| 0| 0/5,2/6,3/5,2/2,0/6,1/5,4/5,0| 0| 0| 0| 0! o > | SS je 5 + i: 2 =f == L L w] 0] 0| 0| 00 01181114 8%, 010,000) |, € 3 1 ’ A > a i c| 0] | 0} 0) 0! 0/4,3/5,4/1,8/2,0/1,6.4,310,5| 0] 0} 0) o| o| [05 DES iJ | ¥| 0] o| 0 010,0! 15650138501 10,9) 0/0) 0 0| o | 38 i og - I | G || 010,1/0,1/0,1 0 1,4 6,1,5,0/4,4/0,114,7/4,4|- 0 |0,1) ea 0 0,1 0,2 Ss 39 ; 1 le Lo) 00101) 0 145,946 wy 5,24,3| 0 (0,2) 0 0,2 0 24 0,2 MI —— — ht — À — + 4} — _ 22 4 — x — —+- Dairy cheese Wi 0! 0] 0] 0| 0 02305 1,8 200,7020,| 0,0 0.010 2163 40 8 P 1 | € 0|0|0| 0| 0! 0 15,25,2/5,2[3,22,5.0,510,20,2 0! 0! 0! 0! 05! 0 | 0 | 3 months lig 0! 0] 0} o| 0| 0 7,2]5,2/6,5|5,616,8 a 0/0/0/0/0 1 | ii do. Fi Iwlo 0/010 0 0/2,012,012,0/1,811,1)0,1).% o| ol 0! o| of 10 mde re bek 1 week | | | iy 1 |} ¥/0| o| o| of 0| 0/5,2/6,5/6,113,6/3,2| 0 021 0 ololokb2l | — == —t - + = I Sun rn =; Dairy cheese tw ol o| of of o| 0 ll 02 Xe 0} 0/0) 0) 0} |. 42 8R i 1059| 10, | 10| 0 1 week | | C2 0] 0| O| 0) 0/4,7/4,512,7]3,8)5,0 0,210.5 02 0} 0) o] Jo, ETE Hvo] oo] 0] 0| 0] bas] bal ol 0] | 0] of 0) | | | Bovine feces? | i | c|0,7/0,510,2 0,5| 0 2,7141 3,813,4 2105220502 0,20,70,5, 0| 10,7 * al c 0,50.20,20,5 0| 0.4,33,4 3,6 34109165 2,9 01 0 110534 Toy] Calf feces 6 i | C| 04,3 02 0 0| 02,52,73,22,01,52,70,9/0,2| 0 | 0 os fi | | 46 | Sourdough | ile ;1 ce JE 010,1 0 [5,8 5,4 4,1/3,7 0,2 0,200,1,0,1| Så 0/0153] | 0,5) | 4.2 | Faa | sr | > i | C0,25,911,0 o 0 0 15,7 5,4 5,238 075 0,201 000/051 104 . Ya i | 150 72 consequently not from dextrose, as has been asserted by LIESENBERG and ZorFrt). Neither do they form any trace of slime with equal parts of the two components of saccharose, dextrose and levulose. Acccording to BEIJERINCK?), the slime consists of dextran. As shown in ZETTNOW’S illustrations (Pl. XXIV), it proceeds from the cell wall. It is for- med more rapidly with 5—-10% cane sugar than with 2%. In liquids with less than 2 9 cane sugar, the slime formation is only slight. Cane sugar broth becomes at first slimy all through, and some few strains keep at this stage for several months (chiefly X- and O- forms), whereas in the case of other strains, the slime soon contracts, so that zoo- gloea masses are formed at the bottom of the flask. On cane sugar agar, the slime deve- lops but poorly, but appears in a very characteristic manner on cane sugar gelatin. Large colonies, clear as water, appear on the plates, resembling the colonies of certain aerogenes species (the slimy aerogenes forms produce, however, slime from all the su- gars which they ferment) and in stabs, we get very characteristic pictures, as shown in the photographs on Pl. XXV. Though these bacteria do not liquefy ordinary ge- latin, and are not provided with other proteolytic qualities, several of them can, after some length of time, liquefy cane sugar gelatin, which figure is indicated in the tables by I (liquare)?). As the genus Betacoccus contains all possible degrees of sliminess and liquefying power, we cannot attach too much importance to these characteristics, and we may find cases where of two strains, otherwise entirely alike (as Nos. 38 and 39, which were, more- over, found in the same sample of material), one will liquefy and the other not. When isolated from vegetable matter, the betacocci thrive as a rule but poorly in milk; when isolated from milk, on the other hand, or from dairy products, and sometimes from dung, they can form comparatively large quantities of acid in milk, and even dissolve some casein (Nos. 29 and 34). The power of souring milk, however, is comparatively soon lost, but can be regenerated by continued transference from milk to milk. The bacteria are often abundantly supplied with lactase, and it may happen that nearly all the lactose of the milk is hydrolysed without any considerable quantity of it being fermented, which shows that the proteins of the milk are a poor source of nitrogen for them. In contrast to the streptococci, they thrive at least as well with yeast extract as with casein peptone … as source of nitrogen. When isolated from beets,they prefer beet juice to casein peptone (Nos. 11 and 12, Pl. XXV). The betacocci exhibit a certain preference for pentoses. Strains isolated from vege- table matter for the most part ferment both xylose and arabinose, whereas those isolated from dung, milk, or dairy products, will as a rule ferment only one of the two, or sometimes no pentoses at all. Of the hexoses, they often prefer lævulose, and of the disaccha- rides, often saccharose. They frequently ferment raffinose, but of true polysaccharides, only a little dextrin at the outside. With regard to salicin, the different strains vary con- ') Beiträge zur Physiologie und.Morphologie niederer Organismen. Leipzig 1893, Heft 1. *) Folia microbiologica 1912, I, Heft 4. 3) ZETTNOW distinguishes between two types of Sc. mesenteroides, Opalanitza and Aller, which, though it is not stated, undoubtledly correspond to the non-liquefying and liquefying betacocei respectively. 73 151 siderably. When they do not ferment raffinose, then in most cases they will not ferment salicin either. The betacocci do not as a rule attack alcohols to any perceptible degree; only a few strains ferment a little mannite (No. 20 even a little sorbite). No other bacteria have proved so variable with regard to the sugars as the betacocci. The power of fermentation generally declines somewhat in course of time, not only as regards lactose, but also for maltose and raffinose; even, indeed for their favourite foods, such as sacchaıose and pentoses, and the source of nitrogen employed often exerts a quite remarkable influence, Yeast extract, for instance furthers the fermentation of raflinose in Nos. 4, 6, and 9, but retards it in No. 33. Similarly, it furthers the fermentation of cane sugar in No. 37, but retards it in Nos. 33 and 35. The betacocci which in a natural state (i. e. freshly isolated from vegetable matter) ferment both pentoses as a rule, however, prefer arabinose, which is best seen when using an inferior source of nitrogen (W); examp- les of this are Nos. 15, 14, 15, and 16. In accordance with this, they are more easily liable to lose the power of fermenting xylose than arabinose, and the A-forms are therefore nearly related to the A + X-forms. It may happen that the A-forms also lose the power of fer- menting arabinose, and of the three strains from the same cheese (Nos. 7, 8, and 9) which were at first entirely alike with regard to their action upon the pentoses, two simultaneously lost the power of fermenting arabinose, while the third, which differs from the others in fermenting salicin, has still retained this power. One of those which lost the power of fer- menting arabinose, on the other hand developed, its power of fermenting xylose, so that ithasreally become transformed from an A-form to an X-form. The explanation is doubtless this that these bacteria in reality possess the power of fermenting both pentoses, and it is more or less a matter of chance which of them they are better able to deal with at the moment (cf. also Nos.6 and 12). For the same reason, also, it is difficult to draw any sharp limit between the true X-forms (which ferment xylose, but never any considerable quantity of arabinose) and the foregoing. The X-forms have only slight slimy growth in cane sugar gelatin, and never liquefy it. They may, however, be found to lose the power of fermenting xylose, becoming at the same time strong slime-formers. By continued re-inoculation of No. 33 in cane sugar gelatin, we succeeded, for instance, in producing a variety which resembled No. 40 in nearly all respects. From this we may conclude that the O-forms are closely related to the X-forms. Without cane sugar, the betacocci form small colonies, and are in all cultural respects indistinguishable from the streptococci. Morphologically also, they resemble the latter (Pl. XXI—XXIV), though the faculty of dividing in two directions is often more developed. The A-and À + X-forms, when cultivated in broth or milk, make shorter or longer chains, and closely related strains may differ in this respect. No. 11 (Pl. XXII), for instance, forms long chains, whereas No. 12 (Pl. XXI) appears chiefly as a diplococcus. When the betacocci are cultivated on gelatin, they form as a rule only short chains. The X- and O- forms will often appear as rods on agar streak (Nos. 33 and 36, PI, XXIII), or as micro- cocci (No.35, Pl. XXIII)}). In fluid substrates also, they can assume this irregular appear- 1) On agar streak, the one (No. 33) of two related strains can form rod-shaped cells, and the other (No. 35) cells of the micrococcus type. We noticed the same thing in the case of Sc. liquefaciens. D, K. D. Vidensk. Selsk. Skr., naturvidensk. og mathem. Afd,, 8, Række, V, 2. 20 152 74 ance, and Nos. 46 (PI. XXIV) and 47 are, from a purely morphological point of view, more like micrococci than many of the true micrococci we have investigated. The above-mentioned morphological differences between the A- and A + X-forms on the one hand, and the X- and O-forms on the other, render it likely that we have here to deal with two distinct species. AS the former always(at any rate unless in a weakened state) fer- ment arabinose, we will term them Betacoccus arabinosaceus, and as the latter (especially the typical X-forms) can be isolated from most cowdung after enrichment in acid sugar broth, we will call them Betacoccus bovis. As the betacocci are for the most part known under the name of Streptococcus mesenteroides,it would have been reasonable to use the name Belacoc- cus mesenteroides for one of the species, had it not been that both comprise strains which do not form slime, and consequently also no mesentery. It is possible that the betacocci should be divided into more than the two mentioned species. With all experiments in this direction, however, I have felt that I was working on treacherous ground, as these bacteria exhibit such great variability in almost all respects. Appendix to Streptococci, Not all chainforming cocci are true lactic acid bacteria, as e. g. the bacterium described by Boekhoul, which in his opinion contributes to the aroma formation in the creamsouring!). Cultivated in milk it inverts a part of the lactose and forms a trace of acetic acid and perhaps also a little non-volatile acid. It is killed at a tem- perature of 57°. Its optimal temperature is 20°, and it already thrives badly at 31°. We succeeded in isolating the same bacterium from commercial starters. It forms no appreciable quantity of acid in milk and does not attack the casein. It is a pro- nounced milk bacterium and grows badly on artificial substrates. In stab culiure on whey agar with 1 °/, Witte peptone it shows distinct growth after one day at 20° but first after 2—3 days at 30°. The chains in youthful state have thick capsules. They can be very long but easily break in pieces. The bacterium is Gram positive and does not develop oxygen with peroxide of hydrogen, which makes it likely, that — in spite of its lack of acid formation — it is related to the lactic acid bacteria. Perhaps it is a variety of Sc. cremoris, that has lost the power to form acid but on the other hand has got the power to form aroma increased. ') Vereeniging tot Exploitatie eener Proefzuivelboerderij te Hoorn. Verslag over det jaar 1917. Al- ready some years before his death V. Storch has found the same as Bockhout, but his work has not yet been published. Micrococci and Sarcine. As already mentioned, there are certain streptococci and betacocci which can appear as typical micrococci or divide in several directions. And again, several of the micrococci can stretch prior to division, or form short chains, thus resembling the foregoing families. We cannot therefore, on morphological grounds alone, distinguish the micrococci from these. In cultural respects, however, the micrococci and the Sarcine are both, as a rule, easily distinguishable by the fact that on gelatin plates, they form larger — often coloured —-colonies, and in agar stab, growth on the surface. This quality also differs in the different species, and in the accompanying tab’e (XXVII a and b), we have arranged our strains with particular regard to their more or less pronounced need of air. The four first (Nos. 1—4), it will be noticed, exhibit no surface growth at all, and thus present no diffe- rence in cultural respects from the streptococci and betacocci, whereas the last (Nos. 30— 31) form a thin wrinkled, mycoderma-like membrane. In contrast to the truer lactic acid bacteria, however, the micrococci and sarcine are furnished with catalase, so that their broth cultures give a development of oxygen with peroxide of hydrogen, — almost always of the nature of an explosion — and this reaction is therefore a reliable charac- ter by which to distinguish them from the other cocci, and especially from the micro- coccus-like betacocci. * As the micrococci on the one hand, in morphological respects pass over by gradual transition into the streptococci and betacocci, so, on the other hand, we find no sharply distinct line of demarcation between the micrococci and sarcine. From our investigations, we have reason to suppose that both these groups can divide in all three directions, but the typical pachet only arises where there is a strong cohesion between the cells after division. In the micrococci, this cohesion is as a rule but slight, and they therefore appear most frequently as diplococci or in a grape-cluster form (as staphylococci) without any pronounced direction for division; in the sarcinæ, on the other hand, it is generally stronger, and we find these, accordingly, far more often than the micrococci, forming tetrads, and now and again even distinct pachets. As micrococci and sarcine exhibit no differences in biological respects, it is altogether unjustifiable to set them up as two distinct genera; the difference between them is not greater than between short- and long-chained strepto- cocci; we therefore suggest the generic name Tetracoccus for both groups. Micrococcus ti in any case meaningless as a generic name, since the micrococci are by no means smaller shan the other spherical bacteria. In the present work, we have only devoted attention to the sugar-fermenting strains. Whether those which do not ferment sugar form another genus, we are not able to deter- 20* 154 76 mine with certainty, as the tetracocci are on the whole but weak acid formers, and their power of forming acid is easily weakened, so that we have all possible transitions between acid formers and non-acid-formers. There is one point, however, which seems to suggest that the two groups should be regarded as two distinct genera, viz. that all acid formers are without exception more or less GRAM-positive, whereas non-acid-formers as a rule are GRAM-negative. WINSLOW even goes so far as to divide all spherical bacteria into two sub-families; the acid formers, or Paracoccaceæ, including my streptococci, betacocci and tetracocci, and the non-acid-formers, or Melacoccaceæ!). | Genus: Tetracoccus (Abbr. Te.). In this genus (Table XXVII) I include all sugar-fermenting micrococei and sarcinæ. From the sugar,they form, besides lactic acid, smaller or greater quantities of acetic acid. The quantity of lactic acid was in many cases so small that we were not able to determine with certainty of what sort it was. As they thus stand at the limit of what we will term lactic acid bacteria, we have not sought for them systematically, as for the cocci already described, and thus make no claim to have found, even approximately, represen- tatives of all species belonging thereto, but merely of some of those most frequently met with in the dairy. As already mentioned, our strains have been arranged according to their need of air (see last column of the table), which I consider to be one of the best characters, even though the surface growth of the stab cultures do not always assume quite the same dimensions, and though a strain which at first exhibits no surface growth at all may in course of years develop some surface growth, as happened in the case of No. 4. The strains which spread most markedly on the surface, however, are found on the other hand to grow poorly deeper down. In shaking cultures, the most aerobic forms grow only on or close under the surface — and this even when the substrate contains sugar. Like the coli and aer- ogenes bacteria, as a rule they spread more on the surface of sugar-free substrates (AG) than on those containing sugar (SG). A small amount of sugar (but not more than 44%) on the other hand will strengthen the colouring. This is, by the way, not a little variable, and it is therefore very unfortunate that most of the bacteria coming under this head have been named thereafter. There are, it is true, pure white forms, which never exhibit any trace of colour, and the typical Aureus-forms also are fairly constant in respect of colouring, though the first generation may be a good deal paler, if sowing be done from the bottom of the stab instead of from the upper part. In most strains, however, the colour fluctuates, from one inoculation to another, between white, greenish, yellowish and brown- ish. Sight alterations in the composition of the nutritive substrate will at once affect the chromogenic power. Quite fresh cultures are often white, only becoming tinged after a few days, or even weeks. , Some strains remain white for years together, and only in a single generation exhibit a slight shading; others may be coloured, and then become suddenly white. At times, the surface culture consists of white and coloured rings, or a coloured star may form in the centre. On sowing out from the white and the coloured ') Systematic relationships of the Coccaceæ. New York 1908. 77 155 portions separately, the first generation will be white and coloured respectively, but in later generations, the two will as a rule again grow more alike. A Citreus-form from Kral, which, like most of the other pure yellow micrococci, does not ferment sugar, was one day found to exhibit yellow centre and white margin; the forms isolated from centre and margin respectively retained their individual colours, pure yellow and pure white, for four years, but were in all other respects entirely alike. In the fifth year, the white variety turned pink. The size of tetracocci is highly variable. Strains which appeared gigantic on isolation had in the course of a few generations come down to the usual size. Spores were not observed in any of our strains. Nor did we succeed, by direct obser- vation, in discerning any motility. Nevertheless, we must suppose that the most aerobic forms (the Albus-forms, the mycodermalike as well as those which do not ferment sugar) are capable of moving from a spot as, in old agar stab cultures, we constantly find isolated colonies under the surface or at several mm. distance from the stab. None of our strains could stand any great quantity of ammonia, and they are there- fore not able to effect any considerable fermentation of urea. The greatest transformation of urea was effected by the Aureus- and Albus-forms, and by the intermediate dung bac- teria. On the other hand, most of our sugar-fermenting strains can reduce nitrate to nitrite, (the non-sugar-fermenting Cifreus-forms lacked this power) thus differing from the strep- to- and betacocci. Though nitrate reduction appears to be a constant quality, it is never- theless not always to be used as a species character. True, it agrees very nicely when we find that neither of the two closely related strains, Nos. 1 and 2, nor either of the two identical strains, Nos. 30 and 31, are able to reduce nitrate, but the agreement is no longer maintained when we come to consider the two strains, Nos. 3 and 4, which are entirely alike save for the single fact that one of them reduces nitrate and the other not. It would be unnatural to count them as distinct species on that account.. The tetracocci usually liquefy AG, though as a rule but slowly, often, indeed, only after several weeks at 20°. The rapidly liquefying forms also liquefy SG, albeit somewhat more slowly. Nos. 30 and 31, which are otherwise slightly liquefying, liquefy SG a little faster than AG. Only strains which liquefy SG strongly (Nos. 9—13) attack the casein of milk to any considerable degree. Exceptions are, however, Nos. 3 and 4, but they are not able to break down the dissolved casein any further. The pathogenic tetracocci (the two Aureus- and Albus-forms) are killed already at 65°; most of our other strains, however, can stand a considerable length of time at 70° or even 75°. The pathogenic forms grow at 45°; but the maximal temperature for the re- maining forms is rarely much above 40°, and often only 37/4”. The optimal temperature is 30°, and the minimal rarely much below 15°, though the growth may still be rapid at ordinary indoor temperature. The tetracocci can stand high concentrations of sugar and salt, though as a rule they thrive better in weaker concentrations. No. 1, which was isolated from anchovy pickle with 15°/, NaCl, grows most rapidly with 5°/, NaCl. The power possessed by the tetracocci of forming acid in sugar solutions with 10—15°/, common salt (See Table XI b) can as a matter of fact be used as a specific character for them. As the tetracocci can stand both heat, and high concentrations in the nutritive substrate, it is not strange that they should u ot) Den + - À | 156 | 78 Table XXVII a. Sg it | 1 | | og ‘ oo | | | so FoR, Heal | Sl al 2 2 2 2 | 17753 Tetraeoeeus |22/°s/ =| 9) S| 212518 2|2|3]2 2121812415 1395 No. isolated ralrol 9 Sl) g£|A E = = a | BE SE 2 3 |z°8 s|22l2) s|s|o| S S 522 from: 228415 |%* 2932 SF FOBIF BREL 238 Sal] VER == A | EU ue | | | | SE — — :— = ; == —+ H 1 | as d | C 103 '0,8)1,910,1/0,50,510,7 0,2 1, 0 pickle i Il ER Bee | _ | Condensed | 5 2 | yenstLextracki 4,7454.,3 011,1] 0136| 515,4| 5,2 Milk from | | 3 | women with) q | C 16 : 4,14,314,1 1,6 Ol! 4145) 34 mastitis | | » 18886002 0 a Dairy cheese’ 5 sp | 1 week 2,715,214,3 Dairy cheese| 1,611,911,6 6 8 R 3 months 5,0,5.2,3,8 7 NE | 0! 012,0 2,0 1,8 1,12,0 24 hours | J ‚1 | | C |1,6 | 0,6,114,34,3 4,310,7 2,9/3,4 Dairy cheese | a 1 T Ze | Ei 8 8 P 9 || W 10,2 ‚20,9 0,9/0,8)0,8)0,9)1,0)1,2/0,2 lweek | | Cc |9,9 v 03,6 3,6) 1,8/2,5)3,4|2,9/3,2/0,5 Micrococcus | | TT 9 casei | à |W 10,5! 0 1,011,4 \0,211,4,1,4| O liquefaciens, | | | | ‘ FEE HE | | C 11,8 7 2,9 3,2 1,8/3,8/3,6,0,7,0,7,0,8 FE Fe | iT Î Dairy cheese, | - | white, 10 8R | a | W 10,1! 0 |0,8/0,8/0,5.0,7)0,5|1,0/1,0) 0 HERE | quite fresh. | feet 0 2,92,92,32,011,413,.213,4| 0 P light IL | | | | brown. | wo 232,316 2,0.0,1/9:3'2:3/9.1| 0! 0) 0 11 Butter 8 | d | 1 | 0 0,1) 0 3,4 71 +. | yellow | cheesesour!) | | « 14 ,714,718,6|, 7 3,6 1,1)0,910,7/0,7|0,9 0,7 + | > pink) | | 0 0 /2,7|2,7| ’ 2,01 00,2] 0 0,2! 0} 0} 0 — — == Dairy cheese Iwi o 9 8 ar 4 2 À ' Ic (3,8 — hours | i i " Mc r | | | | 13 e.pyogenes | | |. 13,01 | | |<- 2,044,5 4,1/2,9 1,6 4,54,1/4,5 + = 25 —4— — —- >= | | | + ees Bel ee Be ee Ee ea ee PS Re ee ER on Le ‚12,32,311,7 1,62,311,91,8 1,0 4,1 | 6,314,7 323,2 3,8 4,313,8 1,8 > 2,3 16 | — || + | orange tak | 313,6! 30, 5 BRE » aureus, Kral | | + 1,84,3 3,4 2,3 2,5 3,8)4,3 3,614 + pale F | | ell ' | | Mc pyogenes | | 14 aureus, i | C \2,0)-- |= | OlufThomsen | mettre 712,5 a 0,5 ” 0 | + orange 1) Cheesesour is a butterfault. 79 157 Table XXVII b. se] | | | | | muk duel, | 28 a | © | o | | 2 | Mi BE SS | Sho Tetracoceus Sels als elslölsiel22 2 232322. 825 _ | fying In ol SZ i BSB 9} 0) & 5123 Sl! || ss 2I2| eis! 5/2 SU El 0 za Cus isolated BSIESl 0 1 =IS | Blais Si sl ss Sl Ol S| 2] we] gs] RY Eu lo af 85) .4°9 No. OSSE | 72 Siva o|s 3 is Sia) 2) S| si Sl S/o] & | al 8) 55 = gl =o, from: aa SEN 2] 2/7|=| SAS) S| 3) | a) | 9) 2) 9155| 88] Total N Las GR 28 eo] | ae Fs\ + = = ee — ——_——————— LORS LE DE = | | | | | | | | | 15 | Condensed 13,612,50,911,613,4.2,712,90,7) 0 0210 0| 3 251167] 5,3 | + | 0 | 0 Sencw Milk 1 tly ay L | | white. FEE TU TTT TTET Yellow | FAG | | ver 16! spotin |; 4,1/2,9)3,6|1,813,612,711,6-+- is | 0! 4120! aa o2|olo!+| » experiment = | FRI | =) | cheese 15 | | LEA an | i— eh: | ; ; — — — - | | | | Fæces 3 1,66,1 16 0 0,5 1,4,0,9 0) 0! 010,5 | 2,0 | | 0 | 0 | + | orange = Pre 18110 0] 011,6] [18] | + | 0 | + |Prown- ++ | + + r T + — 0,5! 00,5 olıal |11| | + | 0 | + | white T T T | | a = | | | 5 a 02 1,81,10,905 0/02 010 09 a LOI tree 1 pet T + 1 ; 1 > | f 0,7 0,7)181,41,6020,7141,614+ | 0111) 118,26, 0 +0 + white —— i — + 17 + + + + + æt à + 0,9181131090,86811,111,20,1 0011011113125 31 010104» Condensed Bea. Le | | white, 1,110,212,5/2 = 10,2, 0| 0 6 | light Milk 2 ‚1105212,5 2,3 al | 2 | og 0 10/014 Bas Tr T = Condensed | | x a i 4 yellow, : 1,112,512,5[2,5 - |-+- |0,5|=- [1,8 ) 0 + | light Milk 2 | 2 | ae A CAR ål hal Ei we A Bi \ "brown Dairy cheese LANL 6|1,4{1,411,1/+ 0,2) 0! | | A 9R 10,7 | | 1,4! 3,9} 091 +10|+ very 14 71 99% | 5 exten- ME 11,6 4,713,6 2,3/2,3| 0 0,7 0 12,5 | aie SS SS Ss mule i Vegetable ‘A margarine 10,911,8.1,8 1,3,0,7, 0 0,2] 010 LEO} Obes 0, E » not keeping | | | | | | it år | I ze | =} tr Surface of I | | | | Cammembert | ? 0,9'2,0 1,81 |= |= |= |= [0,3 (NOR 0 0 |0|0!+ » cheese 2 | | | | 1 7 . + tt | +- 9 | = |: + — 2 | Mc pyogenes | | | me albus 1,10,9/2,7/2,3/2,0/0,9/2,7,2,9/1,6/2,3L2- |1,1/0,7| 0! | 2,0 +/O}+] » OlufThomsen | | bats BEAT, | i dl an a [5 T ICE | 5 LENS ac oe Fu spl cise 2,713,413,213,22,5 2,511,413,6 272,525 O |+- + |2,5| 6/4,1| 96) 16)/+/0/+) » albus Kral | | | | BR) be E SS SS SSS SS il 7771 || | | ii ik = Surface of | fess | | | | | | | | EE > 2,0/1,111,913,2 1,4) 0/1,81,1 09 + le nl | 3,4 | 660,7 | +|+ | 0 derma | | | ike Les) CAM LR SU I ne | ERG Experiment 10|2,7| 3,7) o1l+|+lo| » cheese 27 + 2,7 1,8/3,84,5/2,90,914,1,2,3 3,20,9/+ + + 158 i 80 frequently be met with in condensed milk!) and other extracts concentrated in vacuum (Nos. 15, 23, 24 and 2). As seen from Table X XVII, the less aerobic, as also the mycoderma-like tetracocci form dextro-lactic acid, whereas the lactic acid in Aureus- and Albus-forms, as far as thev produce any noticeable quantity at all, is inactive or lævorotatory. The by-products are as a rule acetic acid. Those bacteria which formed too small a quantity of lactic acid for iden- tification will possibly burn a part of the sugar entirely, to carbonic acid and water. They are, at any rate, like so many other aerobic organisms, capable of burning a quantity of organic acids, as for instance that formed by sterilisation of the sugars, and the degree of acidity will therefore often decrease gradually in cultures with sugars which are not fermented; this is indicated in the tables by —. The decomposition of proteins, however, also contributed in some degree to this diminution of the acid. As the limit between slight fermentation of sugar and none at all is somewhat vague in the case of the tetracocci, the relation of these bacteria to the different sugars has not quite the same value as in the case of the other lactic acid bacteria. Some points are, however, fairly characteristic. The two least aerobic strains, for instance (Nos. 1 and 2), ferment arabinose, mannite and salicin, whereas the two mastitis bacteria (Nos. 3-and 4) do not ferment any of these sugars. Nos. 5, 6 and 7, which do not liquefy gelatin, have all three shown a decrease, in the course of years, of their power to ferment xylose, but all still ferment arabinose, raffinose and salicin. Nos. 9, 10and 11, whichare powerful liquefiers of gelatin, do not, on the otherhand, ferment the mentioned sugars to any considerable degree. The typical Albus-forms (Nos. 28 and 29) are distinguished from the typical Aureus- forms (Nos. 13 and 14) by fermenting sorbite, and often also pentoses, raffinose and salicin. The mycoderma forms (Nos. 30 and 31) likewise ferment pentoses and sorbite, and have arelatively high fermentation of glycerin. They therefore thrive well on the sur- face of cheese Where a splitting up of fat takes place. As regards the morphology of the tetracocci, we have little to add to what has alrea- dy been said. Of the strains investigated by us, only No. 8 (Pl. XXVII) appeared through- out unaltered as a typical sarcina in all respects. And in accordance with this, it has a knotty, or at times even mesenteric surface growth, and forms in broth compact small clumps which at once sink to the bottom. No.3 (Pl. XX VII) has, in broth,a slight tendency to sarcina form. No. 11 (Pl. XXVI), which in milk originally formed pachets, lost this quality through regular re-inoculation, and then appeared as a diplococcus. After long periods of rest (it could be preserved for three years in starch water without re-inocula- tion), it however regained the sarcina form for a time. This feature, together with the strong yellow (even at times orange or pink) markedly spreading surface growth, distin- guishes No. 11 from the related strains Nos. 9 and 10. As I have previously called these important cheese bacteria Micrococcus casei liquefaciens?), they should, according to the altered nomenclature, be called Tetracoccus casei liquefaciens; 1 prefer, however, simply to call them Tetracoccus liquefaciens. Nos. 5, 6 and 7, which are also of frequent occurrence in cheese, can reasonably be called Teiracoccus casei. No. 5 is distinguished from the two 1) We are here concerned, of course, with sugared milk; the unsugared, it need hardly be said is sterilised. *) ORLA-JENXSEN, Doktordisputats 1904. 81 159 others by forming large capsules in milk (Pl. XX VI) which do not disappear again (turn into slime). Nos. 3 and 4 we can call Tc. mastilidis, with the reservation, however, that it may be found to be the same species which produces the so-called staphylococcus-ma- stitis in cows; a point which I have not myself had any opportunity of investigating. The table shows, as distinctly as could be wished, that there are many other points of difference between (Tc.) Mc. pyogenes aureus and (Tc.) Mc. pyogenes albus than the colour alone, and we are therefore perfectly justified in regarding them as two distinct species?). The tetracocci which we have isolated from condensed milk and dung appear to be inter- mediate links between these extreme forms. Nos. 30 and 31 can suitably be called Tetra- coccus mycodermatus. The true lactic acid bacteria as a rule coagulate milk in the deeper layers first; this species, on the contrary, coagulates milk from the surface downwards. 1) The only pyogenic Citreus-form I have investigated is, as mentioned, non-sugar-fermenting, and Gram-negative, and thus belongs to quite another genus. It liquefies AG extremely slowly, and is not able to reduce nitrate to nitrite. A number of saprophytic yellow micrococci behaved in the same way. D, K. D. Videnskab. Selsk.*Skr., naturvidensk. og mathem. Afd., 8. Række, V.2 21 Rod Forms. The rod-shaped lactic acid bacteria are on the whole stronger acid-formers than the spherical, and this is most distinctly apparent when yeast extract is used as source of nitrogen. If they grow at all on gelatin plates, the colonies are as a rule not visible until after the lapse of 5—6 days. They fall into three well characterised genera; Thermobacter- ium, Streptobacterium and Betabacterium. Of these, the thermobacteria, which do not grow at ordinary temperature are more anaerobic than the other lactic acid bacteria, whereas the streptobacteria and the betabacteria behave towards the oxygen of the air — as indeed towards most other things — like the corresponding streptococci and betacocci. The streptobacteria do not develop gas, and always ferment salicin; the betabacteria, on the other hand, develop more or less gas, and never ferment salicin. Finally, to the rod-shaped lactic acid bacteria should be added some small rods, the genus Microbacterium, which, however, differs quite as much from the true lactic acid bacteria as the tetracocci. Among the microbacteria should possibly be reckoned the Bacillus acidophilus often men- tioned in medical works. And in conclusion, we should mention Bacillus bifidus, as being — albeit without further justification — generally reckoned as belonging to the rod- shaped lactic acid bacteria. Genus: Thermobacterium (Abbr. Tbm.). The thermobacteria are as a rule not killed by heating until a temperature of over 75° has been reached, they also require, as their name implies, a high temperature for their development. With the exception of No. 8 (Table XVIII) which grows at 18°, none of the strains investigated develop at anything below 20°—22°, and even at 30° the growth is very slow. Therefore they will not grow in gelatin plates. The optimal temperature lies about 40°, and here the growth proceeds at a furious rate, as long as the strains are not weakened from any cause. The maximal temperature is as a rule 50°; Tmb. bulgaricum (No. 14) grows at 521/,°, and the mash bacteria at even higher temperatures, as is known from the distilleries. The method of enriching thermobacteria in milk or mash is therefore to let the liquids stand at temperatures between 47° and 54°. They may likewise be forced in fresh cheese mass which is strongly heated (cooked). These bacteria are quite extraordinarily particular in respect of the nutritive sub- strate they require, and we have not succeeded in producing artificial substrates which Peo ee 2 83 161 Table XXVIII. | | | el | a aa 17 pot paper nee ay Bgl. a} eo] | CR | | Pa Milk ND . Bsiesiele al E AISI fis! SISU SIS sie 3 3 5 2 1 to! — %n of isolated SEE) 2e s sé) S|sl< 8) 3 3 82 3]8| 803 23 Total N from: Sabre «mæ a1 4,0) = Clg ne | 2 | AE) go ak: II | | | SERGE ‘DN ——— + _ | ER ! RE eee ER —— 1 Sour mash 1) 1 Yo 0| 0; 0/0] 0 ar 0/0] 0/81 00 001010 | 0 | >» aa [mot 0] o| oo o | 7,4] 43 29) 0] 0 |36 o | 010101010 | lo ¥|0| 0! 0} O[0} 0 [14,4/14,4/10,1/ 0 |11,0/10,6 0 | 0 103,402] 0 0 Fry Foto 0/ 0/0| 0 (14,5|14,2/13,1] 0 H6,0/104| 0 | 01023250 | | 0; | |, [clolo3 01,2) 0 112] 2,7] 3,6| 5005| |7,5 27 |011,61,2 0,8] ¥ {0/0 o! 0,9 0 11,1 112,4 13,1112,410,9 /14,9)13,1) 7,4 10,8 0138130) 0 | IR | a lokg 32] 32] 221,1] 0 | 0226) 5 ol gic s , | | | | Milk left to | | 2,01 1 |14,0| 6 stand 24 1 c 0! 0 | 7,0) 7,0! 6,100,2| 0,7| 8,3) 5,0) 5,2) 6,8) 4 10,4 hours at 47° | Hel. | | | Y 0) 0| 0! 0/0 16,8 111,3 9,7 9,210,7 |10,8)12,4) 7,210,1,/5,21,1, 9,5 2 7 | Milk (Milk, | C 10) 0 01010! 0 [11,5] 9,0/11,5/5,6 11,0! 7,9! 9,0! 0,7) 0 32101) 1,1): |, cbr le: Chr. Barthel Y 10 0|0| 0/0} 0 |12,2113 10,614,1 10,8 13,5} 81] 0,5) 0 412,9 1,1) Sr Tg Milk heated | I. SP tar ee Se lo h t | | | | | à BH aed left | reol 0} 9] 010) 0 | 0,5]10,1} 0,7/1,4| 0 | 2,3) 0] of 0} | 11,111 11621. eg to stand 94 | | ¥ 01 | 0) 0/0] 0 |16,2/15,1/13,5)7,9 114,9111,9] 9,0) 0,7) 0 320,5 6,3) 2 11,7) ” | ? hours at 45° | | | Emmental il LA [DR | (af RAT PUR if Bug 9 | cheese 1 | 1) 2/01 005] 010! 0 | 6,3111,3111,09,5 | 0,5) 5,2) 2,5) 0,91 0| | | It) + iy 41196/21,8 . Yiolololololo 12,2/11,3| 11,7 3,4/0,2/3,4/1,1 10,1) ae Burri I L a ic As r Se | ya N re PT POTTER ap # pen ‚le Jo} 0] 0] 0/0] 0 | 0,9] 9,7] 0917.9] 0 los 05 | | 07, {ya 0821264 Y |0| 0] 0| 010! 0 | 9,7110,6110,8 8,11 9,7| 7,4| 2,9]0,5/2,9/0,9| 7,4 ”122,4/18,1 Burri ea Ra LES IL |__| (El he le 1 bus STER - | Emmental A SS, BA 2 520 3h 44] 1 45) o Tol ol oo | 11 | cheese 3 241°) OF * | 0 | 3,2) 2,9) Be 1 13,318,818,9 Freudenreich ol 00|0|01 0 “is 0 (7,9 10,610,1! 8,8) 0,2! 0 1,1! 0 0 | + —= — 1 + - —— — — ——— == Emmental | T T T Î Î 1 | Î | | re cheese 0! 0] 0} 0/10! 0 | 0,7 6,3! 7,783) 0 | 9,9! 0,9} 0 | 010,5, 0 0 | 1 |27,5/36,1'34,6 Bacterium 0| 0| 0} 0|0| 0 | 3,8110,1110,118,8|| 0 |10,3/11,9 0 | 02,3) 0, O | 1 |19,6/16,9,20,0 casei € VER) | | Enalish Sm SS > i + nglis ; 0| 0/0} 0 | 1,1) 54) 254,7) 0| 0 | 68) 01010100 | 11270, 126 13 | Yoghourt | ' | y lo! 0| 0/ 0/0) 0 10,6| 521831 0 | O| 0| 0} 0} 0; Of 0 |.1 21,67 | ” : - (Lactigen) | | | | | TT Dnoitol |. [oil Pal EE: Il - 2 > SÅR 2 | | | Paes OPO ET 27] 1002| 0 1111 ee |. bots rian Ic o| 0107 0/0|.0 | 47 6,5] 2511,11 0 | 0,1) 79] 0} 0| 0} 0,0 | 1 16,4 19,9 16,7 Yoghourt 1 Y0!0|0|0/0|0 | 27321 18 01 01 010|010/0100 | | Il | | r I il i | 162 84 can altogether replace the natural milk or (non-sterilised) mash. For the milk bacteria, whey-yeast-extract-agar, was found best. With the exception of No. 6, (which in contrast to the other thermobacteria grows well with WITTE peptone as nitrogen food, and which we have also kept alive in casein peptone for a couple of years without re-inoculation) they cannot be transferred every time from agar to agar, but often require a passage of milk, or they will be weakened. Nos. 12, 13 and 14 should best be constantly transferred from milk to milk. If chalk be added to the milk and the mixture be shaken up now and again, the bacteria will retain their vitality unimpaired for several months, but when using milk without chalk, it will be best to re-inoculate each week. No. 12 ‘will however, even under these conditions, keep unimpaired for several months at 15° and remain alive for up to two months at 20°, though it will after a time be perceptibly weakened. No. 13, which is probably identical with the yoghurt bacterium first described by BERTRAND and WEISSWEILER!), since it forms similar pennete colonies, I did not succeed at all in transferring from one artificial substrate to another. It only takes on artificial sub- strates when coming directly from the milk’). The lactose-fermenting thermobacteria of mash does not form acid in milk; the maltose-fermenting thermobacteria of milk, on the other hand, thrive well enough in mash. For mash bacteria, a sterilised malt extract solution is not nearly as good a nutritive substrate as might be supposed, owing to the fact that an essential quantity of the nitrogenous substances therein contained are pre- cipitated on sterilisation. An addition of yeast extract renders it more suitable, and we have found that a solution of the trade malt extract (with abt. 50 % maltose and 0,5 % N) to 7 parts of yeast extract (with 0,5 % N) gives a good nutritive substrate, with the sugar concentration most favourable for these bacteria (see Table II d). Stab cultures in high layers of an agar thus prepared are a good form for preservation. The lactic acid formed by thermobacteria is as a rule levo-lactic, more rarely (Nos. 12 and 13) inactive. In addition to lactic acid, they form some acetic acid, and, as BARTHEL first showed, a trace of succinic acid*). In powerful milk cultures of the thermobacteria of milk, there is often so much gas produced that the curd exhibits fine stripes, and the quantity of lactic acid rises to 114 %; for inactive acid indeed, even up to 2%, %, which is far in excess of the amount of lactic acid formed by other lactic acid bacteria. The thermobacteria of milk can under certain circumstances easily become slime- formers. BuRRI and THönı have shown, for instance, that No. 12 as a rule becomes slimy when it has been cultivated for any length of time together with a certain mycoderma species*). The slimy varieties are just as powerful acid formers as the non-slimy ones. We have a slimy variety of No. 13 (presented by Mr. BLICHFELDT, manager of the laboratory of Monsted’s Margarine factory, in England), which, in contrast to the streptococci, so strongly retains its power of forming slime that even after long cultivation at high tem- peratures we did not succeed in transferring it to a non-slimy variety. The thermobac- ') Annales de l’Institut Pasteur 1906, Bd. 20, p. 977. *) We have not, however, tried the malt germ decoction suggested by BERTRAND and DUCHACEK + (Biochemische Zeitschrift 1909, 20. Bd, p. 102). ”) Meddelande Nr. 69 fran Centralanstalten för Jordbruksförsök. Stockholm 1912. ‘) Landwirtschaftliches Jahrbuch der Schweiz, 1909, p. 271. 85 163 teria of milk attack all casein strongly, with direct formation of mono-amino-acids, and they are therefore of importance in the ripening of the cooked sorts of cheese, which are not till 24 hours after cooled below the temperatures most favourable for develop- ment of thermobacteria. No. 12 in particular is of the greatest importance in the ripen- ing of Swiss cheese (Emmental cheese)!). The thermobacteria do not as a rule care for pentoses and alcohols. Only No. 6, which is further distinguished by being able to ferment inulin, exhibits any considerable fermentation of mannite. Most of the mash bacteria do not ferment galactose and lactose, and the two yoghurt bacteria do not ferment maltose. Saccharose is not fermented by Nos. 1, 2, 12, 13 and 14. The maltose-fermenting thermobacteria as a rule also ferment some dextrin, and often also a small quantity of starch, besides raffinose and salicin, Yeast extract will at first retard the growth of the yoghurt bacteria, and can thereby prevent them altogether from attaining development in certain sugars (lactose, for in- stance). There is altogether something capricious in the attitude of the thermobacteria towards the sugars, due to the fact that our artificial substrates did not entirely satisfy their needs?). The thermobacteria (Pl. XXVIII—XXXIV) are pronounced long-rod forms, with a tendency to grow out into threads, often strangely curling. They consist, however, of several segments, which is not to be seen in the water preparation, but is distinctly apparent when the preparation is laidin Canada balsam (See Pl. XXXII). In a young and vigourous state, they occur for the most part singly, or two and two. As they are seriously affected by the oxygen in the air, they assume irregular shapes in streak cultures. Here, they always form an extremely thin layer, and some few (as No. 14, for instance) do not grow on the surface at all. When stained with methylene blue, they generally prove to contain volutin grains. No. 13 does not appear to form grains, and No. 12 but rarely. The granular formation is most marked in No. 14 (Pl. XX XIIJ), which is also in German often called Körnchenbazillus. With some strains, most grains are found in the quite young rods (No. 14 for instance); in other strains, again, in rods 2—3 days old (as No. 6 Pl. X XIX) and the size and shape of the grains depends to a high degree upon the nutri- tive substrate. In the case of No. 14, the mere fact: that the milk used had been heated some few degrees more or less is suffitient to alter the picture entirely. Normally, this yoghurt rod, when stained with methylene blue, has round, dark-blue grains, and no capsule; when cultivated in pasteurised milk, on the other hand (and especially in milk heated for half an hour to 80°) it has oblong red grains and distinct capsule (Pl. XX XIII). According to Gram, Nos. 12 (Pl. XX XI) and 13 (Pl. XXXII) are not stained completely, but exhibit a quantity of irregular granules. Something similar may be observed in the case of other thermobacteria, if very old for instance, or if stained in a too acid state. As these bacteria are such strong acid-formers, it is altogether best first to neutralise the cultures to be used for colour preparations. The thermobacteria are often over 1 thick. In old cultures, they can develop greatly swollen or otherwise involved forms. 1) See further in my Dairy Bacteriology 1916, p. 112. 2) This explains certain points of difference with other writers. BARTHEL (I. c.), for instance, has found that the identical strain No. 12, which we have worked with, ferments mannite and a small quantity of saccharose, which we have never been able to observe. 164 86 The bacteria of mash are noted in the literature under the little characteristic names of Bacillus Delbrücki and Bacillus acidificans longissimus. As they are derived from the mashed grain, I propose to call them Thermobacterium cereale. There may possibly be several species (as for instance lactose-fermenting and non-lactose-fermenting), and I would in this connection refer to HENNEBERG"), who has made a very thorough study of the lactic acid bacteria of mash. The common thermobacteria of milk (Nos. 6—11) should of course be termed Thermobacterium lactis. Here also it is possible that there may be several species (the three first, for instance, which are isolated from milk, grow more strongly in agar than the three last, which ‘were isolated from Emmental cheese). They have all, however, one point in common: like the mash bacteria, they form levo- lactic acid. This is also the case with No. 14, which is an even more pronounced milk-rod than the foregoing, and therefore does not ferment maltose. As ‘we have always found this rod in genuine Bulgarian yoghurt, whether obtained through Professor METSCHNI- KOFF or directly from Professor PRANTSCHOFF of Sofia, I consider myself justified in cal- ling it Thermobacterium bulgaricum. The bacterium of Swiss cheese, No. 12, which we have formerly called Bacterium casei €, I now propose to call Thermobacterium helveticum. Related. to this bacterium is No. 13, from.the fact of its forming inactive lacticacid, from the strong acid formation in milk, and from its morphological features. It differs, however, in the - fact that like No. 14, it does not ferment maltose, and in the far greater difficulty it finds in growing in artificial substrates, as also by its markedly radiating colonies. We consider therefore, that it should be established as a separate species, and as it is said to occur in yoghurt, we can call it Thermobacterium Jugurt?). Appendix to Thermobacteria. We may here mention an interesting lactic acid bacterium which we have come across during our controlling work with commercial starters for creamsouring. Even where these have proved pure by direct investigation they may nevertheless contain a trace of mould, yeast, or rod-shaped lactic acid bacteria, which can gradually develop in the dairy. We therefore always make the additional test of leaving an unopened jar of the culture to stand for a week at 25°, which is the highest ‘temperature used in the souring of the dairy starter or the cream. The detrimental contamination will thus accumulate, while the good acid bacteria (Sc. cremoris) perish in their own acid. It may then chance, that the starter, instead of becoming sterile by this treatment, as it properly should, becomes transformed to a pure culture of a long-rod form, which in milk forms over 2 % inactive lactic acid, and which also attacks casein (12,3 % SN and 13,5 % DN). Even at the optimal temperature, 30°, it does not curdle milk until after 2—3 days, and is inclined to render it slimy. It is sucha pronounced milk bacteria that in artificial substrates, it ferments practically no sugars at all. In older cultures, it is no longer so markedly a long- rod form, but breaks up into chains of short segments, which can be over 1 thick. It 1) W. HENNEBERG: Gärungsbakteriologisches Praktikum. Berlin 1900. *) In Bulgarian, Yoghurt is called simply ,sour milk“ (kisselo mleko) but in Turkish, Jugurt, and it is this word which has passed over, under various forms, into the other languages. 87 165 (Pl. XXXIV) is incompletely stained by the Gram process, in the same way as thermo- bacteria Nos. 12 and 13. It thrives well at ordinary temperature, but does not grow at over 38°. It is thus not a thermobacterium, but in spite of this, it appears most nearly related to those thermobacteria which form inactive lactic acid. As we are not quite certain as to the systematic position of this bacterium, we must refrain from giving it any name at present. On plate XXXIV it is designated as Thermobacterium No. 15. Genus: Streptobacterium (Abbr, Sbm.). These rod forms we have called streptobacteria, from their tendency to chain for- mation. As they grow, even at their optimal temperature, more slowly than the lactic acid cocci, they do not make themselves apparent in: spontaneously souring milk until after the milk has curdled, but they will then, on the other hand, gradually supersede the lactic acid cocci, being able to stand, and capable of forming, up to twice as much acid as the latter. They therefore abound in butter and cheese after more or less time has elapsed. They are likewise always found where vegetable matter is left to sour. There are probably a great number of species, which are difficult to distingusih one from another, owing to the many gradual transitions between. I will here content myself with making distinction between the two extreme groups, the typical cheese bacterium, S{reptobac- lerium casei, represented by the Bacterium casei a which I have formerly described, and the typical vegetable bacterium, Streplobaclerium plantarum. The former we have never met with in vegetable matter, but the latter is of frequent occurrence in milk and dairy products; it cannot escape being introduced into the same from the food and bedding (grass or straw) of the cows. The streptobacteria are always killed by heating to 75°, many strains already at 70°, and some few even at 65°. The optimal temperature is 30°, and the maximal as a rule from 3715°—40°. Only a few strains can grow more or less well at 45° (Sbm. casei Nos. 4, 26, 31, 32, 33 and 34, and Sbm. plantarum Nos. 28, 29, 30, 32 and 34). The minimal temperature lies probably in most cases at about 10°, but the growth here is, even with the most favourable source of nitrogen (Y), so slow that it will often be impossible to discern anything at all until after 14 days. The growth on gelatin plates, also, at ordinary temperature, is very slow, and the colonies are still considerably smaller than those of the cocci. They grow better, of course, on SG than on AG. Streptobacterium easei (Table XXIX) forms either pure dextro-lactic acid or more or less considerable quantities of lævo-lactic acid in addition, so that we also find inactive lactic acid in the cultures. The power of forming dextro-lactic acid is, however, by far the most constant, and many strains which at first formed almost exclusively inactive lactic acid have yet in the course of years ended by forming pure dextro-lactic acid"). As men- 1) We have consequently been led to investigate considerably more strains than we otherwise should have done, for of course we could not know that two bacteria, otherwise alike — even though isolated from the same place — were identical, despite the fact of their forming different lactic acids. These parallel investigations of identical strains have, however, further increased our knowledge of the same, and it is interesting to see that they often exhibit exactly the same mutations at the same time. They may, however, also be found to differ suddenly in their relation to one or other of the sugars. 89 167 Table XXIX a. eS aa elec |: Te (ada To À Streptobac- rey Sec a= © 2 2 2 2! à | = 2 2 | 2 | o vw | : = | 2 c lene «SE = No, terium casei Ses 3%) 8 FIR: E Slal3| 2/82 sia|s|sig =|5|'2| SSB Go| % of isolated Sts Se) >| Rial a) 5| & Sy eS We ee | S|] 2| | S| 5 |e5| 33]Total N. from: Er 22° BEN => = Slå” S 1217 AP] @ 185] 55 — ; À | | ol ASS ISRIDN — —— = = =: — = = 1 = — = = + i" | ig 1,1} 2 | wW| 0| 010,2! 0! 0/0 | 2,5! 2,5) 2,6] 1,81 0 | "1010101010 Mazun ; Yale | 1022 1 | püggeti d |c| 021,4] 0| 0 N 3,2| 4,5] 4,1| 4,51 0 | 0,5! 1,410 | 0 0| 0 0,7) | Pal 92 8,8 7 ’ | Y|| 0] 0| 0} o 00,9 7,7 7,9| 7,4] 4,5] 0 | 3,6) 5,610 a 0,0 Ba i 4 =! É 4 —t —! | ie ae _— Buena i Ww 010,10,5| 0! 0 [0,21 4,1) 4,2) 2,7] 4,1} 0,1! 0,2) 2910 || 0} 0| 0 7 11 7 | na le 0 0,511,4| 0 |0,2/1,1/10,1/11,9| 9,2] 9,2! 0,9) 0,9| 7,0/0,5)0,5)0,2) 0! 1,6) 43) 7.0) 9 197 1 Re af Ol 6,1] 114,4 | 4,1 0 | 0| 0| 0} 23 Sere eet 1,6 | | ; |W] Of 0105| 0| 0 0,1 3,8) 4,5) 3,8) 3,4) 0,2) 3,6) 4,510 | 0| 0 o| 3 122 3 » a c| 0! 00,9 0, 010,2 11,5112,4113,7 10,4) 0,2) 6,5]10,8/0,5}0,2|0,2) 0 | 7,0111 | 9,7 17,1119,4 | Y 4 | 5,9] 14,01. | 2,7 00 1,8! 0 10,6 ! A — | | L KE [TE T BER 0 | Butter 6 Le iy W |p, 51022/0,5| 0| 0 1,9) 5,0 5,4] 52] 4,6) 0,2 3,2) 4,110 0 panes Me EA ( tose ) C |0,5/0,9/1,4/0,2/0,2/3,4//11,0/11,0/11,0/10,1/ 1,1) 4,7| 8,8,0,5/0,710,7) 0 | 5,0 AS di a Ir! ey 63 [1401 | | 0 jo | 0]2,0] 0! 92 — [Dairy cheese i W |0,4/0,2/0,5| 0 | 0 12,51 5,6) 6,1! 2,9) 4,1) 0,5) 6,1 af 0] 0; 0 |3 | 99 5 . ” I > ee ae a? | €)05| 00,5, 0} 0 5,4)12,8.13,3) 30106 09126! 05/0 | 01021 0 74110! 74 ee ee * 18 561 12,6 7126| 0 |0 | 011,6) 0| 83] | FØR: = || A! = Ls H 4 I ae I — — Dairy es) i : | 3,6 | | 2 “A ia |wlo2| 0102| 0| 011,61 6,3) 5,6) 4,1| 4,5) 0,3 3,6 0101010! |6 126,1, he || dG) | c 1050235 0| 0 2,3]11,710,4 8,6104 0,7) 5,9) 23010702) 0| 65 aa a ae Di ain LI — ———— (—H— - +- 0,9 Dairy shiceed W} 0) 0/ 0} 0) 0 0 4,5 4,4! 4,5) 4,1) 0,1) 0,2) 4,1/0 | 0| 0 ol 2 7| or | % | 61-0) 0! 0) | 0/1810, 12,8110,1/10,1) 07 0,9! 9202 01 0) 0! 61 3 111,3] 9,5) 8,0 aa 3 months tO Y Jo 0.0 my 0152] 146 | 05, G7] 101010101886) | | a a |W j0,2| 0| 0| 0 0/18 43 4,3| 43) 380,5 321 4310 0] 0/0] [11] 7,4 11,7113,0 ; | (2 010 o| o| 02,3[11,3,11,7/11,9| 9,9) 0,2] 5,6| 9,910,20,210,210,2] 5,6] 3 12,2) |“ Dairy cheese i i 21 43 i i d i , , 9 9 P d W 0 0 0 0,7 5,9 5,6 6,1 5,0! 1,5 4,7 6,3 0 0 0 7 10,4 8,4 3,4 | 3 months _ | € lo,70,20,9,0,200,2/2,9/12,8/12,2/12,6/106! 5,0! 8,3/11,310,5/0,7/0,5| 0 | 9,2 4 | 471 Dairy cheese wlolo À 0! 012,0) 6,1] 6,3) 5,2 wal 0,21 3,4 5,910 | 0! 1,6 1 6 P d 2 (11,5 4 9 Lo c lol 0! 0| 0| 0/23] 8,3110,1| 7,7110,8| 0,6! 5,6 9,9, 0 | 0] 010! 6,1 2 Zn ae | Y|o|o/0|0|0[59 14,6) 0 0 | 0} 0| | 62 au RON le ils u te | | Emmental w02 Bi o|olı,l 52| 6,8! 5,4) 3,6) 0,2! 1,1 50010|0|0| | cheese d | | 101 | | $ 047963 en c lol 0! o| 0! 0|2,5]11,7/13,7|13,5|12,2) „',12,4113,5|0 | 0102| 0 10,8, 7 111,5 ,7/26,; asei a. ae | | | | N Y |0,7 7,1115,8115,1114,6113,8] 0,2| 0,715,3)0 | 0| 0| 0|12,8 | | | | D. K. D, Videnskab. Selsk. Skr., naturvidensk. og mathem. Afd., 8 Række, V. 2. 168 Table XXIX b. 2 VÆR | | | See p=S [sal£le 2 2] 9/2] 3 g\2| 3 2 o|ol2l.lels erıum casel Sos © | © © E a 3! =) = Eu = = = 2 8 als = 2 No. : Spo Weer lisa] Be) eho] ae] el El a) SNS PS lElS isolated Ox |3E| | 2175| REA 1285 [71 58] PG EN EST = = Ets kai kl S| slalsl 8 81S! sia) ==] Si sisisia | from: | a" || | alm! | Ike AS a | Le 1 — Ee | | Milk | wo! 0 0! 0) 0 19 j2 | Weigmann’s a | ! | collection c |0,2| 0 0 No. 32 ae — T Dairy chees | W | 0 10,210,2] O 13 I7B d | 3 months 2 010 0 0 vi ai |wl10102 0 ) d | € 10,7.0,211,610,2 at, ‘Dairy chees re id 0m 0) = I cha“ nz | HR pr i W || 0| 0/0,1) 0 16 » | di | d C| 0|0|0|0| 013,6 | Y || 0} O| oo! 0! 016,1 == 11 — on Dairy Rs | i |W10,50,5| 0} 0| 0/18 1701 di | he Se a | c] 0] 0] 0] 0| 0129 | 06,1 | IL ALS BI ERDE [61] | IW| 0| 0} 0} 0} 0/ O 4 di {| | | 18 | » | a | Cc} 0} 0} 0)0,2| 012] | | y| o|o| of 0| 015,9 Dairy cheese] wos o| 0} 0! 010,2 19 | 8 P lied C 0,5 0 0! 0) 03,4 3 months | Y 10,9 °| 0/0068 i al = | wi olo! oo! 024 20 ERE (Wer Pi Må | | 921 | CIS LD ER 0| 00,5] i . 1 { 70 21| Kefir2 | d | co2 00,70 70736 | EEE 2.2 he BR RT irv cl Il | | 99 ‘gee 5 | a |wlozozb2! 0} 0123] 7,0 2 D onfhe a Eis 1,40,2 0,2 5,0 nn ta ae Co 3,8 6) | | | ff UT dk: + |W} 0 10,210,3! 0! 0 2,0) 4,7| 5,4 3, sa! aa] oz 0 23 » d | | | k i C 0,5 0! 0| 0! 0 2,5/10,2] 9,7) 7,9) 9 9,0) 8,6 1,902 20 91 169 Table XXIX c. - - . - Streptobac- | „23 |< | a 2 à o| o! 9! 0] of 2 | o|| | We MUR No, terium casei Eich gh) 8 2 SE = =j3|2|8 £ 2 8 2|8 £ Elsa Eee em] 0, of Mpotea 282 ÈS PARIS) El ES PSS) Sl4 |) |S El Bl S| 5 en 28/Total N. from : Hee CE Ps) ee 7 I À Aas t ea) "ir ja ||” HE £ | | | A | | IF © SNDN A | | eal | | | | 2 ee oo 4 |W 02031 0| 0 #4 5,4| 5,4 5,4| 5,0) 0,9| 5,41 4,3 ol 0 0,0! | 3157140 71138 | | | | HÆL FE TOT Be | [¢ 0,7, 0, 0} 0} 0/2,3110,811,011,311,5] 3,6 9,2) 8,110,510,2 0,0! 6,5) à] “en = == + ———— tt —+—— | + t + i | wo ologl o| o [A] 5,2! 59! 521 4,11 16! 36 46 010/010) | 2h. |. 25 | d 0 { Ole steaks malt? 0,5! ” |1,6| | cae nl a ho ol 54) 65 | Cc 0,9 051,6 0,20,2 34124 11,9 12,4/10,6) 4,7) 6,511,01,10,7 0,7, 0 9,5 : Swedish | | Fe a PANNE | | | manor-seat | | | El ‘ciieods a | € 0,910,710,710,50,514,3 10,4/11,0/12,2 8,1) 3,4 3,210,60,50,90,70,2 34! 5 104! 51, 3,2 TRES | Y 18 0! 0! 0! 017,0] 114,914,9113,1| 8,8) 4,5/12,60,20,20,20,2| 3,8 ; curvatum | | | | | Misc DR 1 is ii Po Ft Dairy cheese} i | 16) | 5,6 | ihe ||| | | |W 0,50,710,2| 0 | ? | = ai ER 9510,70, 10 0: 1,8) 5,9! 5,9| 5,0] 4,1 34 5,6 #501] 0 ad 72 6,7 4 weeks d re 0,7 1,1| 0 11,410,7 3,4 13,1112,4 12,2/12,4 4,1 10,1 10,8 0,2! 0 |2,5| 0! 7,2| | | i | wo aml 0 0 DEE 50 481 521501 1,8 501341010) | | | glıoal | 28 d | c 021,4 0| 0 14,515,0112,2114,0112,6111,5| 4;1| 9,9/11,50,510,512,302 9,5119 | 8740186 3 months | Y Da] 0| 0 5,915,9 153 5,0 0! 0120| 0 11,0! | ” Milk left to | hss Wall | gg (stand at 30°] 4 | C |0,2/0,9| 0| 0 3413,2/10,1| 7,7/11,0110,8| 1,8] 2,9! 9,000,2) 0 | 0 0| 7,2 a ha! sal ga with 1,4 Jo LY 3,8)6,] 11,7 | 3,8] 6,1 0 || 0 10,210,2| 9,9 SUS REESE lactic acid | | | | | | | | - i = za! Zu er ET = = 2! : 30 | Faces 2 d | c |0,710,910,710,93,212,7\11,3110,4 9,2] 7,4) 3,8] 4,3, 9,0/0,7/0,7|0,9/0,5| 8,81 3 /15,3| 4,8) 4,7 > A Ik a a ee | EEE 31 Mes d | C |0,90,510,9 4,1)3,4/3,6)10,4) 9,7, 9,9 DE 2,3, 8,60,710,711,1/0,5| 6,3) 3 114,4 AT + I or LE = eh la ie ET EEE 39 | Butter 1 a | W 0,50,5/1,6/1,4/2,0 7 7,0| 7,0! 7,0| 6,1! 2,3] 0,3) 6,51 0] of | 2 116,09 los 5 fine | | € lo,710,9/2,7 2 14,4/14,0/14,2/13,3| 4,7 1,8 13,1/0,20,7/0,5| 0 7,9) 3 14,6 vw 1 1 EI = = EE LO — Emmental | Led eg eae (a | | | Ww ? 3,81 7,2 31 | 3,6) 6, 0 | | 33 |cheese from) dc 0501105] 17025] 68) 721 70 Sl os] SO Bahr | | 5|15,3)/18,7}22,0 „Alnarp“ | | “ I1,111,1/2,5[5,012,314,5114,0115,3114,4113,5| 2.9) 4,7113,5/0,510,910,9| 0 | 8,1 | = i + ts a - Hr Cheese Er . 8 05 | | 9 | | | | 34 er d | €/1,10,911,615,014,115,9116,4 14,4 16,0113,7| 6,8) 6,1114,91,40,7/0,9 0! 8.1) 9 |16,0128,0119,7 a | Y 0,71,1 PRE 2 15,1 | 6,5| 4,7/12,2/.01 0| 3| 0| 7,9) N | | | | | | | Pag tioned in the introduction, both the source of carbon and the source of nitrogen can, in this species, affect the relation between the different lactic acids. And we may note, as a curious fact, that Nos. 33 and 34, which only form dextro-lactic acid of lactose and all monosaccharides, also form inactive lactic acid from saccharose and maltose. Streptobacterium casei thrives well in milk, and can, like Thermobacterium lactis, form abt. 114 % lactic acid therein. In accordance with its slower growth, however, it takes 22: 170 + 92 at least a couple of days, and as a rule from 3—5, or indeed often longer, to coagulate the milk. It always'attacks casein, though with varying strength, in the same manner as the thermobacteria of milk, and is therefore of the greatest importance in the ripening of cheese. Some few strains can for a time, though for no perceptible reason, render the milk more or less slimy. One such slimy strain, of No. 34 (Pl. XL), lost the power of fer- menting cane sugar, though its power remained quite unimpaired as regards all other sugars. Later, when its power of forming slime had disappeared, it at once regained the faculty of fermenting cane sugar. Streptobacterium casei for the most part ferments galactose nearly as strongly as the other monosaccharides, and lactose more strongly than the other disaccharides. Never- theless we may find some few strains altmost or entirely losing the power of fermenting lactose in artificial substrates (Nos. 5 and 23). Yeast extract will in many cases increase the power of fermenting saccharose (Nos. 2, 3, 5, 12, 15, 17, 18, 19 and 26). A similar effect is produced by this extract with regard to mannite, which is fermented — albeit often only to a slight degree — by all streptobacteria. Nos. 27—34 ferment sorbite and most of them also rhamnose, a little inosite and dulcite. As they, more- over, form only dextro-lactic acid, and exhibit lively growth at 45°, they resemble Streptococcus glycerinaceus in almost all respects. Sbm. casei has no marked power of fermen- ting pentoses or polysaccharides. Now and again we may discern a slight tendency to fermentation of arabinose, more rarely xylose. No. 16, in the first years, fermented some inulin, but lost this power after a time (its power of fermenting cane sugar decreasing at the same time). i Streptobacterium casei (Pl. XXXV— XL) forms, in broth, chains of short rods, with the ends as a rule cut off straight. The chains have often sharp breaks, are as a rule very long, and tangled, and then flake off, while the surrounding liquid is clear. The rods can be so short that if rounded, they may resemble streptococci (No. 9, Pl. XXXVII and No. 28, Pl. XX XVIII). They may, however, also grow out into longer rods, often curved. In milk, the chains are as a rule shorter than in broth, and on solid substrates as a rule even shorter still, but on the other hand, we often find longer rods here. On agar (No. 34, Pl. XL nethermost) and at times also in broth (No. 2, Pl. XX XV) screwed chains are often discerned; these arise by bending of the rods themselves. Even quite short rods can be curved (Nos. 33, Pl. XX XIX, as here the layer of indian ink is too thick the bacteria appear too thin) and unite two and two with the concave side inwards, forming rings or shapes of a deceptive likeness to that of the micrococci. These irregular shapes are most frequently met with in the rhamnose and sorbite-fermenting strains which, besides the qualities already mentioned (growth at 45° for instance) also differ from the remaining strains by exhibiting livelier growth with 2 % common salt than with only a trace of the same. There is thus much to advocate separation of these strains as a distinct species, which might suitably be called Streptobacterium curvatum!), if it were 1) Miss TroıLı PETERSSON (now Mrs. ALMQUIST) has, in her “Studien über die Mikroorganismen des Schwedischen Güterkäses” (C. f. Bakt II. Abt. 1904, Bd. XI, p. 137) already established a certain species of lactic acid bacteria, Bacterium curvatum, distinguished by just these above-mentioned morpho- logical qualities, One strain of this (No. 26) kindly furnished by the writer in question, exhibited lively growth at 45°, but did not otherwise agree with my typical strains of Streptobaclerium curvatum. The extremely slight splitting up of casein (only 3.2 °lo DN) might, pike bck seem to suggest that it has degenerated on the whole during the long time of preservation. 93 171 Table XXXa. ri fe. ey de al Ciel twine ik has sol | | 9 | ae Milk Streptobac- ZEIGE S| » APE 2 2 218 o | 9 2 MCE |e es terium ABS 5;8) = 5385 3 8/3 8 82 82 8153 38 Shee. plantarum 23/32! >| 0 5 =s|s is 5 Sıis/ 213 a| =| Si] €) 8/8) 31.353 isolated from: Se 3°] |” ler tele E | © a| = a) | Cr ER os i j | | 2-4 EEE < | | | | | | =: Tam if ——— ; | i i | | | w| 001 25! 0| 0! 0! 3,8) 5,2) 47 3,8| 5 5,6! 50! 501 o | | | 0,5 | Butter 2, d 101 187 | T| 6,9 fine | 451050,70,7| 9,7/10,1/10,1! 7,7)10,6, 9,9) 7,2'10,1) 0,54,31,4 9,9) | 34 ’ | | |» | | | | | 3,8| 0| 0| 0 15,5113,5 14,2 10,1/15,8/14,6/14,9/15,8| 0 4,31,4) 9,9 Mik | |, 1,8 | 1,8 47 | 0] 0] 0} 2°} 4,1! 3,6] 3,6| 2,5) 72 > ( 2 | Weigmann’s | i CRE #12 0 A 0 x] M | | °| = PA collection No. 3 2,3) 0| 0 |3,3111,0110,6112,2) 7,9| 7,2110,6| 7,7| 0 | 0 0,2 0 8,8 f 3 | Butter 7 Ty : ; 1,6 stot 0} 4,1] 5,2) 4,1) 2,31 0 | 4,5] 4,1] 0 | 0 02 Ol 113134 EFT: | cheesesour | 6,1! 0! 02,3) 9,9111,0110,8| 7,7| 7,410,4| 9,0! 0,5 0,70,9 0! 8,8 2,5 al Butter4 |, 1,6| 0} 0} 0) 4,1! 521 36 2,31 o | 4,8! 36 Of o 08 0) 13/34 | cheese-sour 6,110,2| O 11,1 9,5[11,7/11,5 8,1] 9,9110,8 9,5) 0 | Chat 0| 83 2,0 > == F 4 ACER ES r Fe Ait PAIE A LOT, i A ae omg 1,2| 0| 00,8 3,4! 3,4! 29 23) 0,1 4,5) 4,31 0 | 0,102 0 4h13l.08 ; 7,4) 0| 0 13,410,6 9,7110,6| 9,5) 0,7/10,1/10,8) 0,2) 0,709 0 8,8) | 3,2 4 weeks | g ? å Dai = = je et ct x Î r t fe + js He T | À oe É SEE i 0,7| 0! 011,1) 3,2) 3,4) 2,9! 3,2] 0,1! 4,3! 3,8) 0 | 0 0,2 0 | 4 90 tek 8,3| 0| 0 11,812.2112,812,210,11 0 111,711,01 0 | 0,20,9| 0! 8,3 wee zat a gg! P| Sana LER 3 I 9 Lt || N » 0,1, 0) 0/9) 6,3) 6,3) 4,7 4,1| 2 | 3,2) 50 0 | 0 | 4| 7,2 7 10,6 4,1 | 4 weeks 0,90,213,641111,911,7110,8| 8,811,910,4 9,51 0,9) 790902 9,9) ‘ IB A LE SRE RO LEE HEN uk BERN ae ZRH 9,3] 1,1] | | | ialw 01! 0} 0/61 6,5] 741 4,7) 3.6) 2" 2,7) 52) 011 0 8 ) di b | | def > d 1,410,2/2,5/3,812,2/11,5|11,7) 7,7/11,7' 4,5) 9,7 045/1% 0,7 ol 9a] | 29708 0| 0/5,6/5,6 9,0 92 000 81 | | | s pee pl nr i ae | | asd el L | 1 02 of 0/14) 50] 5,2) 41! 3,61 41 591411 02] 0 | | | 9 » i 1,4) 0| 0 3,911,3111,011,0 9,7) 9,9 11,010,8) 8,6) 0,50,7 0,5 9,0, 4 9,0 0| 0} 0/7,2) 122 10,8 10,4) ala) | ”| | ? | | | | wa » i 00/009 45 4,1| 4,3 2,5) = 5,6) 38.021 0090 25 | ante 0,9, 0! 0 3,410,811,5111,9 8,8110,111,010,4 0,5) 0,71,1 0| 9,0! | 3,2 Dairy cheese ol ol oltel al 42! 50l 20) 02 asl aaloalo! | | | | a 2P i pad ME PAR A 14| 3,6 3 months 1,1) 0| 0 '2,9/10,8'11,0 11,3) 7,3) 8,6|10,6 8,8) 2,3 ER 0 84 [18 = ED Tree ERA | ED ÆDE tre allie = 0| 0| 00,3) 3,6 3,6! 3,8) 2,51 16] 3,4) 3,8 ol 0 12 | Weigmann’s | | | 0,2 | 10| 6,0 collection No. a 2,0| 0 0 3,4 10,8 11,0 11,5 9,0) 7,7,11,9110,4| 9,5, 0,5 0,9 0! 9,0) 4,1) | | || 172 94 Table XXX b. fa Streptobac- [23 a Bale als 2 | 2 - erium Solo S| & dl? LOL es 2 | © (9) plantarum TE 28 3 $ 5 a = | = 8|8 2 2 8 © | orl eS Mi - = ee no a en sage] | 5 |£ EEE EEE 8) sl elelsl ay Et alzl |= SEES ER FE HS IS al eles SE 4 2” | lass 8/2/23) Saas 9] |_|} a oa Al@| A jos! 3% o of = Ar 100 a een ER SIE ge Total N h 2 | EE Eu AES panels wl 0107 01010 il T ge | | 8/49) SN |DN collecti i i 52} «| 5,0! 4,7) 3,8) 2,0 4,3 i ere c 0,9 2,7| 1,6| 0 1,43,2 1 en LO des ep o. 27 ly 05 01010 DE de 13,1/13,5/13,3| 9,9} 9,7112,4| 9,7/10,6 4 (11,3 Dairy cheese | En 131 9,9 “hr ur å we 8| 81) 39 40 TUR: | W ‚6 2 0,8 des i 010501] 1807| 4,1) 41) 3,8 1,8 +- +4 , ; nths ch, naar Poele | ‚1 1,4| 1,1/2,911,813,2]11,9 4,5) |” 0,5 Dairy cheese | + 8 3,2/11,9/11,5/12,4) 9,9/11,0/11,7 4 110,6 AL nae? ,1110,6110,4 ’ 2,9| 2,6 9R i | 10,2) 0,5 2 502118 Nå | + 7 | 4,1] 4,1 3 months CG 1,1 1,8 10,6 3 4 + a ee 3 4,3 2,3 0,5 4,7 3,0 0,5 i | u 2 Keñr6 li c 09 0,9 se el PAL Bi KR 7110, 41) 1,1/+0,6 JL 2 ? | Calf fæces 6 | i | C 10,9) 0,7 m 5 = [Ord 8,3) 8,6) 6,11 6,81 9,1 1,5) 8,6 | Î ‚9 0,7] 8,8/0,5)2,5)2,7) 9,7 5.9| 8 eee ee nn 7,41 210,4 ; 2 6| 5,6] 6,5 zer 4 1,0! 2,4 i i | G 10,9! 0,7 051 ‘ 2 2 2 ,5| 8,1) 5,6| 7,0 ; y 4 7 > 1 3 4 3 6 2 ? Camembert- | . | W 0,1 01! 1 A al ol 9,9 8,1 9,5 6,8 8,6 9,5) TT 83 6,3 10 5,2 3,3 + 0,1 ch 1 | be rl 2,5 5 ES 5,9| 5 Bee sel ESEL Ban = le 0,5) 0,2) 2,9] 1,81 0,1 a > Be! ETS © Dairy cheese | , | i 4] 74] 7,7| 6,3) 3,6) 8,1) 7,2 56 2,9 6P i\w 10,2! 0,2| 0,5 a 6,1 ‚9 1,0+1,0 Ask a Le lool 05! 11 1,8/2,0| 4,5| 5,0) 4,6| 3,8] 5 i + ' I ths > 0,5 i 5,0 3,8 3,8 11 5 11 9 12 ? så 2,9 3,6 0,1 Dairy cheese | MAT j i 1/10,6)12,6) 5,9) 9,0) 4,1 : ee 16,3) 18,7 5R w 10,1) 0 | 0,5 | , LD) A ; 4 4 months c 10,5! 05 x 02 Fi 1,6 4,7 5,6 4,3 ÆT 5,2 3,8 47| 0 | A u Bere & À o , sg | ‚| 1,1/0,2/2,713,4/12,8/12,8/11,9/11,9/13,3) 7 314,6 .|w 10,1! 0,2! 0,5/0 == ‚3 7,7/11,3) 0,5 log 1957 17,3 potatoes di 1) 0,2! 0,5)0,1)1,4)1,5) 5,3 lente 2 Du 3 1 11 | G 0,8 6,6 0,3 0.5142 31 14 ee 2,8 3,6 0,7 5,6 18 01 LER | A| 9}12,7|10,1 Satser ) | W 1 »912,7,10,1112,7111,0111 D) 1 II |d(i) À VE 0,7 1,010,1[1,6|1,8| 6,1| 4 IE ER: 2 39, 6,7 3 [15,3 4,4 5,9 née = he | Dr 8 6,8 10,5 0,4 3,6 7,1 12.5 12.2 ’ 9 4,3 4,1 3,5 0,5 | rr a | Wot] 2,5) 380,1] olo,7 | An ne 2155| 51 5,9 : | MARGE 8 2102/02/29 26) 19] 14) 10 22] 69 18 01 | 5) 51) 5 pal cen | rt | 24 À j K | ) | wo | = = a? 5,6| 6,2) 7,4| 6,5 ke eg Fe 0,5| 0,70.111,711,7| 6,4 4,2] 5,2] 1,8 PAS 29 0| 0 Ur | | | „7 7,3 0,4 0,4 2 3,3 11 7 119 1 tå ; 155 23 3,8 2,7 26) » 2m laa) 02 5] 060,1! 0 [1,1 à a BE BER un BE: 23 7 — 4 LE alé |C 0,7 6,4 030,20,327 ef hae 3,6| 3,4] 0,2) 4,5] 3,5] 2,5 MERS 15 217 ali) c 103] 61 0303033, ds DEE 63 5,3 —— tt - ’ i — 7 » 211 |d(i)| C Io] 0,1 nae = = i Bs 10,4, 8,2) 32] 8,3 7 1,0 | af 5,4 = 2 ’ 1 It » ımlailc oz 0, ‚0,33,512,2113,9111,9 9,7] 2,9] 9,7/10,4 28] 1,1 pri 0,9 0,7 0,714,2 3,5 10 8 11 1 10 ? N ? 0,4 2 — ———— ‚8111,1/10,6,10,5|| 3,6| 9,3] 7,0! 0,6 21 1,2 DO ane 95 173 Table XXX c. | ea | | | | LL NON lsh we n Streptobae- SES al ,| 2/2] 2/0) 3| 8) 3] 8) 2| 2/2! 8 EA SEN mak | tertum [2e/3ai5| 2) 8] 2|=/31 3/2) 8] 218188 213188 él RE No. sie) 2) pla Blélal 2) 3] a|.8)/4/3/c\ si] Sik] st = 182 eo "lo of plantarum selles) SISSI @) S S| = 3 SS 5 = 35 SI] 2%) Total N. isolated from: = 2” |? | <| A eh Ons | = = ER! gs == | ET | | | Æ3 4S SN |DN : ET T = — ect + + + i + = + i + Î + H + N Soured w| 1,111,4 | | 0,2 | 0| 5,4 Mil | 30| diffusion | i | C 0,73,01,80,21,13,4111,3 13,3 | 0,7 |13,7/0,5| 0/68) 3 122 5,6 | 4,8 slices | LY 10,91,10,5 0 6.1168 13,1 14,014,0 11,9) 914,6 529,2 01154 0/0 9,0) — — —— == —t— — + \ a Soured beet |, W 0,40,60,50,21,311,5 5,6 4,5 5,2 3,4] 3,6! 6,4/3,2/ 0,1! 5,1/2,2| 0/46] Hee hic | slices I | | © 0,6/1,00,5 0,6 1,62,9 10,1 9,0 52 95] 93) 7,788 0,5 | 9,7 2,510.5 4,5, asad ese Lice —— == 7 + an 1 === tt i 4 | FA 4 © 0380,10,50,1)1,3:1,4) : 5.2 5,0| 5,1| 3,61 3,11 47 0|01| 4,7/1,4| 0/3,3| ra 9 | © |0,510,7/0,7 0,4'2,3)2,3)10,4 10,8) 9,9) 7,9104| 8,60, 0,2 11,01,80,2 3,8 2 + 1, |w 5.8| 4,0] 4,6 2,7 32) 4,120 05| 0,111,4 0,24] FA “| 2 Ic 17113116 8,2) 7,7, 9265591062605 45) *| ” al L w 4,2) 2,3 2,0 0,7) 10 601,702 0,204 0.09 Ra Fra C | 81) 83] 80 46 5,3 843,72 | 02 20031681 |" | 1 4 = = 35 I W 4,1 Lo! 0,9! 1,6! 0,72 014! 0 08! 0 0,7 | SÅ | | 3 i c| | 7,3| 6,6| 4,7| 4,5) 4,7 6,3 5.7) 5,9 | 0,11,90,2 391 |” | = = = i tt 1 36 | PURE |! Cc | 6,8 7,5 6,5 5,2 6,0 630205 5| 0105 0 [541 |% i Ze tt a ad de 37 Pickled w 0,7) 1,0) 0,61 0,5) 07 0.610101. 0,102 011] ur | cabbages 2 c | 8,1 2 9,9) 4,3) 8,8) 816,5, 8,1 772007 si] |” a] „|, Iw 0,716 08 04| 1,6 071,5/05 | 0102 rapt), leg 5 | z Cc 92 7,9 8,6 Ld) 8,3) 8,6/7,0 83 | 811,605 7,9) Nas | aa ee Hi 4,0} 2,4| 1,8| 1,7 18 38140;1| | 0104 0 20| 4 Ea L 3/3,6) 8,6| 8,5| 8,3) 8,1) 7,7 9,986 7,9] 0,51,80,7 81] |” | ©] Maire Me ER ka (1,1) 42] 1,4] 1,8 06] 0,9 ENS 0,104 0/23 14 | 40 | ib € | 0,7/3,310,7/3,4) 83| 7:1| 8,3) 6,1) 7,9 8,88,6 5,5) 05200568 |” DÅ AM (0,4! 2,8 0,8| 0,7 os! 0,4) 0,6/0;6 0,1) 0,10,4 010,7 Lil 61 | k 52,9, 8,1, 7,01 88 5,2 6,3 8565 5902180581 | | 8 Ae A 00,6 5,3 7 2,5| 1,7] 1,3 3,535 152] 0,110,4) 00,814] 38 eee TC 27107100105 65 7,212,48,3 92 022902192] "| ” | 23 eee | = T 1 | = i r ee 251 D. |W 120. AILS 4,3) 1,6 1,7 1,1) 0,7| 1041101 0102 Berl lee | . Mic 0,80,7 894,024 2,6 10,7 107105 7,1) 8,9| 9,918,2| 7,6 | 0,2 2,5 0,2 8,3 ! 4 + + t 44 > shale 0,1/0,2 ou 0,1201 9,5110,1 9,6 ol 74! 97] olo1| 01,9 0195 0 | not that we find exactly similar curved rods in the species Sbm. plantarum (No. 1, SG-plate PI. XLI). On SG, the cells of Sbm. casei are most frequently short, and on AG, they are always short, and even rounded, so their chains resemble streptococci (No. 4, PL XXXV, No. 9, Pl. XX XVII and No. 32, Pl. XX XIX). When stained with methylene blue, some few rods (especially in older milk cultures) can at times exhibit dark grains similar to those noted in the thermobacteria. 174 96 Streptobaeterium plantarum (Table XXX a,b and c) most frequently forms pure inactive lactic acid, but can also form dextro-lactic acid; in some few cases, indeed (Nos. 11), 21, 31 and 32) even exclusively dextro-lactic acid. Though, hke most strains isolated from pota- toes, it forms much acid in milk, a powerful splitting of casein is nevertheless exceptional, and has only been demonstrated in the two inulin-fermenting strains (Nos. 20 and 21) which were isolated from cheese?). Nos. 2, 3 and 4 are slimy (ropy) in sugar agar, but not in milk. Nos. 7 and 8 have at times shown some slime formation, especially in cane sugar agar. Radiations from the agar stab are now and then seen, always in the case of No. 11. : As regards its relation to the sugars, Streptobacterium plantarum is affected to an extremely high degree by the source of nitrogen employed. It is common, for instance to find the fermentation of raffinose and inulin — and even of saccharose, mannite and pentoses — fail with W as source of nitrogen. There are indeed some strains which can hardly ferment monosaccharides (Nos. 24, 35 and the sour cabbage bacteria) unless coaxed to a certain degree with their favourite dishes. With Cas source of nitrogen it prefers as a rule maltose and cane sugar to lactose, and often ferments raffinose, some strains also inulin. The fermentation of sorbite and rhamnose is far more frequent than in Sbm. case). À number of strains show a comparatively powerful fermentation of ara- binose, and a few (as for instance most of those isolated from potatoes) of xylose. No. 1 can ferment a little starch. Briefly then, these bacteria can on the whole — as was to be expected of plant bacteria — utilise a far greater number of carbon sources than bacteria living normally in milk, where there is no other source of carbon beyond lactose. From a morphological point of vie'w, it will hardly be possible in all cases to decide whether we are dealing with a strain of Sbm. casei or of Sbm. plantarum (Pl. XLI—XLV). In broth, Sbm. plantarum as a rule forms shorter chains, or even isolated long rods. If they exceptionally form chains of small segments, then these latter are rounded, and consequently much like streptococci in appearance (No. 44, Pl. XLV). On solid sub- strates, isolated rods, or a very few together will most frequently be found. 1) No. 1 has, however, formed a little inactive lactic acid from lævulose and saccharose. 2) We have therefore doubted, if we should reckon these two strains to Sbm. casei or to Sbm. plantarum, and only their powerful saccharose- and inulin-fermentation decided us for the latter. 3) Several sorbite-fermenting strains of Sbm. plantarum (as for instance Nos. 19, 20 and 33) also ferment some inosite, but only a trace of dulcite. Genus: Betabacterium (Abbr. Bbm.). The betabacteria (Table XXXI) are in most respects so closely allied to the beta- cocci that they may be regarded as the analogous rod forms, and they can,„like the beta- cocci, be divided into those which ferment arabinose (Nos. 1—20) and those which do not (Nos. 21—33). The Bacterium casei 7 and 6) formerly described by me are typical representatives of these groups, and as they also differ in morphological respects, — the arabinose-fermenting rods being generally shorter than the others — we consider ourselves justified in establishing the species Belabaclerium breve and Belabacterium lon- gum. Besides occurring in vegetable matter, they are also found in cheese, feces and kefir grains. The rods forming the tissue of the kefir grains (Nos. 1 and 2) are, however, so different from the remaining betabacteria that they must undoubtedly be reckoned as a distinct species which we will call Betabacterium caucasicum, as the kefir rods are now for the most part known under the name of Baclerium caucasicum. This last name is also erroneously used for streptobacteria, which are likewise also found in kefir (though not as a rule in the grains themselves), and which are considerably easier to obtain in pure cultures. Most of the betabacteria are not altogether killed by heating until 75°. They form inactive lactic acid, at times with a surplus of dextro-lactic acid. The great majority of strains develop gas (carbonic acid and more or less hydrogen). When the development of gas is strong, succinic acid is also formed. The power to form any considerable quantity of gas, and thus succinic acid, is, however, soon lost in artificial nutritive substrates, and in most of our strains, gas development could only be observed by sowing out closely in sugar agar tubes. From levulose, some few strains can form a small amount of man- nite, and as the mannite-forming mash bacterium Lactobacillus fermentatum?), which has been closely studied by JAN Smit, has the greatest resemblance to the betabacterium (judging from its fermentation of sugar, with Sbm. longum) it should doubtless be reckoned under this head. Possibly the mannite bacteria*) so dreaded in the making of wine also belong here. Some of these form, like the betacocci, slime from cane sugar, which the beta- bacteria investigated by us never do. The betabacteria grow poorly in milk, and do not as a rule attack casein at all. When cultivated at optimal temperature, however, some few strains, in a freshly isolated state, were able to curdle milk, and even with a slight development of gas, but this power was soon lost. The frequent occurrence of these bacteria in cheese is due to the fact that they are better able to utilise lactate of lime as a source of carbon than are most other lactic acid bacteria. A characteristic feature in the betabacteria is their lack of ability to ferment salicin ‚and alcohols, and, with the exception of the Bbm. longum forms, their preference for 1) Studien über die flüchtigen Fettsäuren im Kase etc. Centralblatt f. Bakteriologie II. Abt. 1904, XII, p. 604. 2) Zeitschrift für Gärungsphysiologie 1915, Bd. V, p. 273. 3) In MÜLLER-THURGAN and OSTERWALDER’S big work: “Die Bakterien im Wein und Obstwein und die dadurch verursachten Veränderungen” (Centralblatt f. Bakt. II. Abt. 1913, Bd. 36) there is also a list of works on this subject. z 9: D. K. D. Vidensk. Selsk. Skr., naturvidensk. og mathem. Afd., 8. Række, V. 2. 23 176 ! 98 Table XXXI a. resis ssl o| 2| 2 g} 2/2] 813i elel 2 aS Betabacterium S355 Big so] 5 mile 5 2| 2) 2/2) s|3| S| 2 Ez] "Io of isolated from 85225525 >| R | à 232151219 8 2 5 33 Total N. Hesse lois) Mi | Rs S| 81213128 Bu un | «|= Si Lei El Lo E = 2°| SN| DN = nt: Kefir 5 0c} 0} 0, 4,2) 0 L i € 0 111,7 Kefir 3 i 0,2110,4 0 14,2! 0 iF Tl eat Dairy cheese W|0! 0!7,90 9 R EEE 0,5 1 week C| 0! 0113,7| 0 D in tenes | | | |] FF airy cheese 2 7R 1 RA KERES 0 4 months | C pa 010,1! 010 Dairy cheese i (d) is 0| 0O| 5,6) 01 010 5 P i | Tlclo1l 01110! 0 | 4 months yo 0, 7,4| 0 Dairy cheese 2,5 5 R ica} + | WO 8,3 0 4 months C | 0! 0,516,4, 0 | Dairy cheese w 4,3) 6 P reales 0| 9} gg 4 months C | 0} 05167 0 0 = — + il + + + W | 0! 0,7| 1,4 0 » Es +e] ol 7al 8,6| 0 al Ay sae 6,8| 7,9) 0 Calf fæces 6 i |+|C 10,5! 0,2) 2,3)0,2)0,2 =} —— = - ‚IK 7 4 ++ — == —t Emmental w olo cheese 0:34) 2,9 (Bacterium En ke 0,2113,5116,2| 0 casei 7) 4 0| 6,8} 9,0) 0 Sour doûgh E | 0 | C 10,1) 0,5 5602 = = - - = mi 12 | » E d | 0|C 10,1! 0.2} 0,210,2 | _ = ue z= _ Feces 3 + || C 10,1! 0,9 3,802 1 1 — = | at To i | 000 10102 200 5 qua on | a | | | } | i 10 C | 0) 11 1,8005) B z | Soured potatoes | mi | i | 0} C | 0! 6,5) 7,7|0,3/0,2/0,815,6 | 11 | BEL | ey tn > 99 177 Table XXXI b. | | Fe, © F | ihe II . EME ale x ele | Milk mr alo ,1° & & A 2 212 22,7 8 2 o 2 a | —— = No. Betabacterium S38 548% 5| 2 a 2 =| 3 = 8 8 = E gis JE 5 2 STE . ‘lo of isolated from Sasse &| 2 ss sa 8) S| S| 8| SI S Sle S Sl ggg Total N. | + FSae lel ed < 2° ke 14/98) l6)s él" | [É3 | Milk LA! 1,6) 1,411,7| 0,9] 0,1 Wei | [EE ARN Je 0,7 3,8| 3,6/3,4 2,0 1,4 | Microbactadaial | 3 rp er : w! 1,81 2,0 20 15 = 4 Months c| 5,2) 5,2 5,0 2,7 3,6 4 Cheese-sour | al" 2,0| 3,011,9 1,5) 0 | butter 6 | C | 5,2) 5,214,7| 1,8| 2,9 j + + — Mik a "| lwl 1,8| 1818 0! 0! 02 5 | hour to inan lc! 1,6! 4,7| 5,2 5,2 1,8 0| 34 agar. tube | i | | | Bac. acidophitus| 4 two 1,8) 1,81,8 1,5| 0,1] 1,6 6 Kral [9° ’G 4,5| 4,7 al 1,8] 2,9) .2,9 7| Motten | Telos j 233216 0129 | G jv, | + »3| 3,2/1,6) 0) 2,9) 1,8 red beet 4 a fof ; = Dairy cheese lw 0 I 1,0) 4,3) 5,0:4.5) 0,9 00 8 6 P da 1,8| 2,02,3| 0,2 4 months | | clos 10,4) 8,3/7,0| 0,2] 0,2) 0,1 | 3,414,1 |W 4,5] |" 0! 0 9 Good butter 1 | "| 2,0/1,1 Ga |clo 5] 97| 6,1! 0| 2,0] 0! 0 ] re 1 | | 10 | Calf fæces 6 li | cl: Ol 1,6] 1611! ol 1,6] 0 EE — = = = =: Bacterium | | bifidum | 11 | Infant fæces MI) d Y-/ 0 52146 1,1/11,5) TIRE: if ly | 0 8,613,5| 0| 9,5 Propionic acid | | bacteria | Swedish manor- | | | 13 seat cheese | C /1,1/0,2/15,3) 0| 0; 0,2) 6,3) 5,413,6| 9,9) 7,2) 7,4) 0,9) 2,3) 0 9,7 +0,8 Troili Pet MATIERE | roili Petersson i | | | dd — N = +. L = =. Emmental choc) | | | | | | | ‘ 14 | Bacterium acidi | C 8,4] 0) 7,9/8,6,8,8 LE 26404 9,7 "T 9,1/12,2 ke 011,436 113183) 3,8 |+-3,7 propionici x ite. | | | gee merne 103 181 - lene blue, it sometimes appears highly granulated, and thus resembles an extremely small streptococcus. Closely related to Mbm. laclicum are apparently the rods Nos, 1 and 2, though they are about normal thickness, and altogether without surface growth’). In streak cultures, No. 1 (Pl. XLIX) is often swollen up to a club shape, and partly Gram negative, They do not reduce nitrate to nitrite, but No. 2 splits up peroxide of hydrogen, Of the strains of Mbm. lacticum above mentioned, No. 6 should be a typical repre- sentative of the acidophile intestinal bacterium Bacillus acidophilus, discovered by Moro?) and FINKELSTEIN’). None of the many writers who have studied this Gram-positive rod have interested themselves in its biology, but have devoted all the more attention to its morphological features, and it has been described, now as a thin form, now as a thick one, now as isolated long rods and again as chains of short segments. As some in- vestigators have observed that it can be very irregular (compare with our No. 1), they go so far as to declare it identical with Bacillus bifidus. This is an excellent instance of what can be attained by morphology alone in bacteriology. As it is quite impossible to determine, from extant literature, what Gram-positive intestinal bacterium is really meant by Bacillus acidophilus, there is certainly no reason to retain the name. When endeavouring to discover the acidophile feces bacteria, by enrichment prior to sowing, in sugar broth with %—1 % acetic or lactic acid, we have either encountered betabac- teria (strains 13, 14, 15, 29 and 30 were isolated in this way) or microbacteria most nearly approaching strain No. 1. Microbacterium mesentericum. This form is, as mentioned, easily recognisable by its enormous mesentery. It is killed by heating to 70°. It is only a weak acid former, and in its relation to the sugars most resembles Mbm. lacticum. It forms long, thin, often highly granulated rods, which, as they lie in slime, appear much thicker in Indian ink preparations than in colour preparations. On AG, where it thrives better than on SG, it looks like a streptococcus (PI. L). Mibrobacterium flavum is nearly as resistant to heating as Mbm. lacticum, and like the latter, thrives poorly at anything over 35°. It also grows very slowly below 20°, and has thus only a slight temperature interval for its development. In contrast to the other microbacteria, it will still thrive in a sugar solution containing 10 % common salt. It can form up to 1 % lactic acid from levulose, but it is only with monosaccharides that it forms any considerable quantity of acid at all. It forms a finely flaked precipitate in broth, and is even more aerobic than Mbm. lacticum. The yellow colour is more strongly apparent on SG than AG. Mbm. flavum (PI. L) forms clumsy rods, measuring for the most part 0.5 x 1—2y, though they can also grow out into threads 104 long. When coloured with methylene blue, it is distinctly granulated. Rods such as strain No. 10 (PI. L) are frequent: in calves’ dung. Treatment with methylene blue will as a rule only stain the poles, and they are thus easily confused with small micrococci. Although the surface growth is more often white than yellow, and they are only weak acid formers, they are doubtless closely allied to Mbm. flavum. 1) No. 2, however, has now, after 10 years’ cultivation, begun to exhibit a slight yellowish-green surface growth, like Nos. 3—6. 2) Wiener Klin. Wochenschrift 1900, No. 5. 3) Deutsche med. Wochenschr. 1900, p. 263. 182 ; 104 2 : heen : ae In milk, and especially in. cheese, we encounter some small Gram-positive rods, which entirely resemble Mbm. lacticum, and can, like the latter, stand heating to 80°, and only grow in the upper part of the agar tube. In stab cultures, the surface growth is a faint yellowish-green, which it can also be, at times, in Mbm. lacticum. They form, however, too little acid to be included among the lactic acid bacteria. They curdle milk in the course of a week without altering the reaction to any considerable degree, and they dissolve the casein gradually. In milk to which chalk has been added, some few strains can form up to 80 % SN, 54% DN and 6 % AN. They therefore also liquefy gelatin, especially that without sugar (AG). Dextrose gelatin on the other hand, will not liquefy if closely sown, as the acid formed will impede the action of the proteolytic enzymes. They split up hydrogen peroxide, but do not reduce nitrates. We ‘will for the present call this bacterium Microbacterium liquefaciens (Pl. L). It is possible, however, that it may belong to quite another place in the system. Bacterium bifidum (Bacillus bifidus). At the commencement of the century, Tıssier!) made the discovery that the over- whelming majority of the bacteria in the feces of breast children did not consist, as ‘with adults, of Gram-negative coli bacteria, but of a Gram-positive, irregular rod form, which could be club-shaped, or even forked, wherefore he termed it Bacillus bifidus (Pl. LI). As the bacterium does not form spores, it will be correct to alter the name to Bacterium bifidum. It is immotile, obligatorily anaerobic, and requires sugar for its development. It forms an acid not precisely defined, but no gas. Casein and gall appear, strangely enough, to impede its growth. It grows for the most part so slowly, that even at the optimal temperature, 37°, it may often take a whole week before the colonies become visible. It is killed by heating to 60°, can still grow at ordinary indoor temperature, and at 40°, albeit extremely slowly. According to Tıssıer’s investigations, it is said not to be quite so predominant in the faces of infants reared from the bottle. We have also investigated a great number of faeces samples both from healthy breast- reared children and healthy bottle-fed infants, the material being procured, partly from private acquaintances, partly through the courtesy of Professor LEOPOLD MEYER, from the Lying-in Hospital. The direct microscopic examination revealed, besides a quantity of more or less GRAM-positive cocci?), on the whole not more Gram-positive than GRAM- negative rods. No constant difference in the flora of fæces from breast childern and feeces of bottle-fed children could be discerned. It was very difficult to isolate the Gram-positive rods, and when we finally succeeded, they perished so rapidly that as a rule we had not time to investigate them further. We were never able to keep a strain alive for more than a year. Strains which could be further inoculated from one anaerobic agar tube (high tubes, from which the oxygen of the air was excluded by means of pyrogallate of potassium, according to SrriBoLt’s method) to another, often refused to grow in freshly sterilised liquids covered with fluid paraffin ') Recherches sur la flore intestinale des nourrisons. Paris 1900, *) In a single instance, we found only streptococci in the fæces of a breast child. It was stated that both nurse and child were healthy. re 105 É 183 (as in the case of the club-shaped strain shown in the photo). Yeast extract is undoubtledly the best source of nitrogen, and with this, we were able, in the case of the two strains shown in the table, to study its relations to the different sugars. Of these strains, No. 11 (the strain from infant III, Pl. LI) is highly forked whereas No. 12 is only very slightly irregular. They behave towards the sugars (Table XXXII) as Bbm. breve and Bbm. longum respectively, and thus probably represent different species. Baclerium bifidum has, however, probably no connection with the betabacteria, and if it were not that we had previously met with strains which, at any rate in a freshly isolated state, formed great quantities of by-products, we should hardly have reckoned Bact. bifidum among the true lactic acid bacteria, as, though it does form some dextro-lactic acid, 30—40 % of the sugar fermented is turned into acetic acid. Unfortunately, we did not succeed in preserving any of the isolated strains long enough to enable us to observe whether the quantity of by-products gradually decreased. The different species of Bact. bifidum doubtless constitute a separate genus, possibly forming a connecting link between the lactic acid bacteria and the propionic acid bacteria first isolated by FREUDENREICH and myselft). The rod forms coming under this head can, as will be seen from the accompanying photographs (PI. LI), also be more or less forked. We cannot here enter further into the qualities of the propionic acid bacteria, and I must therefore refer to my Dairy Bacteriology. In Table XXXII will be found, by way of example, details of two strains in their relation to the sugars, I would merely add here, that these strains are also very strong fermenters of inosite. 1) Landwirtschaftliches Jahrbuch der Schweiz. 1906, p. 320. D. K. D. Videnskab. Selsk. Skr., naturvidensk. og mathem. Afd., 8 Række, V. 2. 24 184 106 Key to Identificatio The lactic acid bacteria ferment carbohydrates and higher alcohols to lactic acid. They do not gro acids therein contained. They are Gram-positive, immotile, sporeless rod- or sphere-forms. We cannot here re nearest related forms. Rod Forms. A. Without eatalase, red a. Produce only a trace Genus: Thermobacterium. Form laevo- or inactive lactic acid. Except Tbm. cereale they strongly break down | casein and thrive well in yeast extract. They never ferment pentoses and frequently not salicin. Long-rods, which grow at 50° or more, but do not on the other hand grow at lower than 22°. Tbm. helveticum. Inactive lactic acid. Ferments maltose, but not saccharose. Able to form more than 2,7 °/o lactic acid in milk. | | Jugurt. Inactive lactic acid. Does not ferment either saccharose or maltose. Able to form more than 2,7 °l lactic acid in milk. » bulgaricum. Laevo-lactic acid. Does not ferment either saccharose or maltose. Forms at most 1,7 “/o lactic acid in milk. lactis. Laevo-lactic acid, Ferments both saccharose and maltose. Forms at most 1,7 "lo lactic acid in milk. i : cereale. Laevo-lactic acid. Ferments as a rule both saccharose and maltose. Does not curdle milk. Genus: Streptobacterium. Form inactive or dextro-lactic acid. Thrive well in yeast extract and as a rule also in milk, Always ferment maltose and salicin and only exceptionally not lactose. Shorter or longer chains of shorter or longer links. Maximum temperature as a rule 35°—40°. Sbm. plantarum. Mostly inactive lactic acid and no breaking down of casein. As a rule prefers maltose and saccharose to lactose and frequently ferments raffinose, inulin and pentoses. Sbm. casei. Mostly dextro-lactic acid and breaking down of casein. Prefers as a rule lactose (especially in milk) to saccharose and maltose and only exceptionally ferments raffinose, inulin and pentoses. b. As a rule produce perceivable amo Genus: Betabacterium. Almost always form inactive lactic acid. Thrive well in yeast extract, but as a rule badly in milk. Never ferment considerable amounts of mannite, inulin, dextrin, starch or salicin. Have a comparatively small mannose fermentation. Bbm. caucasicum. Prefers arabinose. Ferments no disaccharides out of the Kefir grain. » breve. Ferments arabinose with predilection, frequently xylose too. Maximum temperature 38°. longum. Never ferments arabinose, but frequently xylose, and as a rule raffinose. Maximum temperature 45°. B. As a rule catalase, Gexvus: Microbacterium. The undermentioned strains form dextro-lactic acid. Thrive badly in yeast extract. Very small rods, only exceptionally more than 0,5 » thick. Mbm. laclicum. Ferments maltose and frequently some starch. Slight, often greenish yellow surface growth. mesentericum. _ Ferments maltose, raffinose and starch. Strongly curled, white surface growth. 2 flavum. Does not ferment any di- or polysaccharides. Vigorous yellow surface growth. 107 185 Lactic Acid Bacteria. nia salts or single amino acids, but like the animals require complete proteins or the entire complex of amino caracterises the separate species, but-only note the qualities by which they are best distinguished from the = Sphere Forms. ate and surface growth. ets besides the laetie acid. Genus: Streptococcus. Always form dextro-lactic acid and thrive well in milk but not as well or even badly in yeast extract. Sc. pyogenes, some strains of which do not generally ferment lactose, however thrives badly in milk. As a rule they only divide in one plane. a. Mostly shorter or longer chains. Never pentose fermentation. Sc. thermophilus. Ferments saccharose, but not maltose, dextrin and salicin. Does not break down casein Maximum temperature 45°—50°. » cremoris. Does not ferment saccharose, maltose or dextrin, frequently not salicin. As a rule it breaks down casein. Maximum temperature 35°—38°. ». mastitidis. Ferments saccharose, maltose, dextrin, starch and salicin. Does not break down casein. » pyogenes. Ferments saccharose, maltose, dextrin and salicin, frequently starch too. Does not curdle milk. f. Diplococci as well as longer chains. Mostly pentose fermentation. Always ferment mal- tose, dextrin and salicin, as a rule also saccharose. Maximum temperature 45°. Sc. liquefaciens. Ferments sorbite and glycerin. Breaks down casein and liquefies gelatin. » glycerinaceus Ferments sorbite and glycerin. Does not break down casein nor liquefy gelatin. » inulinaceus. Ferments raffinose and inulin, frequently starch and xylose. Does not break down casein. » bovis. As a rule ferments raffinose, inulin, starch and arabinose. Breaks down casein. Does not grow at lower than 22°. 7. Mostly diplococci. Always ferment maltose, dextrin and salicin, mostly pentoses too. Sc. fecium. Always ferments arabinose, as a rule saccharose too, and frequently raffinose and rhamnose. Does not break down casein. Maximum temperature 50°. |» lactis. Never ferments saccharose, raffinose nor rhamnose. Frequently breaks down casein. Maximum temperature as a rule 38°—40°. id other by-products besides the lactie acid. | Genus: Betacoceus. Form laevo-lactic acid or exceptionally inactive lactic acid. Mostly form ‘slime from saccha- rose. Thrive well in yeast extract, but only now and then well in milk. Never ferment inulin and starch, exceptionally dextrin. On the other hand frequently ferment raffinose. If raffinose fermentation fails, salicin fermentation as a rule fails too. Low optimum temperature. _ Be. arabinosaceus. Ferments arabinose with predilection, frequently xylose too. Diplococci or chains. | » bovis. Never ferments arabinose, but frequently xylose. Often irregular forms dividing in two planes. > and surface growth. Genus: Tetraeoceus. The undermentioned strains form dextro-lactic acid. Division in two or three planes. Te. casei. Ferments arabinose, raffinose and salicin. Does not break down casein nor liquefy gelatin. » liquefaciens. Does not ferment arabinose, raffinose and salicin. Breaks down casein and liquefies gelatin. » mycodermatus. Ferments arabinose, glycerin, sorbite and mannite, but not raffinose and salicin. Does not break down casein, but liquefies gelatin. Mycoderma resembling surface growth. 24* The Lactic Acid Bacteria, Arranged According to Their Habitat. The species are arranged after the frequency of their occurrence. 1. Mash, at temp. over 50°. Tbm. cereale. 2. Sour Dough. Sbm. plantarum, Bbm. breve and Be. bovis (dividing in two directions). 3. Souring Potatoes.') At first, chiefly Bc. bovis, Bc. arabinosaceus, Sc. fæcium, Bbm. breve and Bbm. longum. Later, chiefly Sbm. plantarum (xylose-fermenting) and betabacteria. 4. Sour Cabbage. At first, chiefly Bc. arabinosaceus, more rarely Be. bovis. Later, chiefly Sbm. plantarum. 5. Souring Beets and Diffusion Slices. Be. arabinosaceus, Sbm. plantarum, Bbm. longum (and Mbm. mesentericum). (In rotten beets we find, besides mould fungi and pectin-fermenting plectridia, aero- genes bacteria and Be. arabinosaceus). 6. Human Feces. Microbacteria, Bacl. bifidum, Sc. fecium, betabacteria, Sc. glycerinaceus (and Sc. liquefaciens). The great bulk of the intestinal flora of sucking infants consists of true lactic acid bacteria, whereas in adult human beings, as in adult animals, it is, as we know, com- posed of coli and butyric acid bacteria. ly The mass investigated was a mixture of raw and boiled potatoes which had been stored in cold (1) or warm (2) state. The investigation was made partly a couple of months after storage (I), and partly when it was taken out of the pits in March (Il). The figures quoted are used in the tables under the finding place of the separate strains. 109 187 7. Cowdung. Tetracocci, Sc. bovis, the streptococci not fully identified in Table XXIII, Be. bovis, Sc. inulinaceus, microbacteria, Bc. arabinosaceus and betabacteria. 8. Calves’ Dung. Microbacteria, Sbm. plantarum, Be. bovis and Sc. fæcium. In cowdung, tetracocci of various colours predominate; in calves’ dung, on the other hand, rod-shaped lactic acid bacteria predominate. 9. Cows’ Milk, Fresh: Tetracocci (besides cocci not fermenting sugar) and Sc. liquefaciens. Standing at ordinary temperature: At commencement of souring: Sc. lactis, Sc. cremoris, Sc. inulinaceus and Be. bovis. Later: Sbm. casei and Sbm. plantarum. Standing at 40°—50°: At commencement of souring: Sc. [thermophilus and Sc. fæcium. Later: Tbm. lactis, rarely other thermobacteria. . Pasteurised at low temperature: Sc. thermophilus, Sc. fæcium and Mbm. lac- licum. Pasteurised at 70°-85°: Chiefly Mbm. laclicum (besides spore-formers). Yoghurt: Tbm. bulgaricum, Sc. thermophilus, Sc. fæcium and Thm. Jugurt. Kefir-Grains: Bbm. caucasicum (besides yeast). In Kefir also Sc. lactis, Bc. arabinosaceus, Sbm. casei and Sbm. plantarum. Condensed milk and sugar-containing salt pickle: Tetracocci. 10, Butter from Soured Cream. Fresh: Sc. cremoris and Sc. lactis. Later: Sbm. casei, Sbm. plantarum, Tetracocci (yeast and mould). 11. Hard Rennet Cheeses. Not highly heated (as for instance Danish dairy cheese). Fresh: Sc. lactis, Sc. cremoris, Tc. liquefaciens, betacocci and Sc. liquefaciens. Later: Sbm. casei, Sbm. plantarum, Bbm. breve, Sc. glycerinaceus, Tc. casei and microbacteria also gelatin-liquefying forms), more rarely the species occuring in fresh cheese. Highly heated cheese (e.g. Emmental cheese). Fresh: Sc. thermophilus, Tc. liquefaciens and Tbm. helvelicum. Later: Thermobacteria, Sbm. casei, betabacteria, streptococci and tetracocci. On surface of cheese: Sc. glycerinaceus, Tc. mycodermatus, besides other tetra- cocci and cocci not fermenting sugar (also Bact. limburgensis, mould, yeast, etc.). In the case of Danish dairy cheese, no considerable difference was found between the flora in cheese from pasteurised and cheese from raw milk, which is simply explained by the fact that these cheeses are given so much buttermilk that it is the bacteria of the latter which chiefly dominate there. > “et Ries 4 5 2 Experiments with a view to utilising the isolated species. It would be natural, of course, to test our lactic acid bacteria in practice at the places from where they were isolated. There is no doubt, for instance, that the loss of dry matter in souring of potatoes, beets and other fodder could be reduced considerably if the gas developing bacteria there abounding could be at once subdued by rich inoculation with such lactic acid bacteria as thrive well in the vegetable matter concerned. Experiments on a large scale in this direction were also planned by me in co-operation with the State Plant Cultivation Committee (,,Planteavlsudvalg’’) but were relinquished for the time being owing to scarcity of fodder. We have attained some extremely favourable results with sour cabbage, at the Ama preserving factory. Cabbage inoculated with Bc. arabinosaceus No. 19 and Sbm. plantarum No. 37 developed a much better aroma and was more tender than cabbage which had soured spontaneously. Under the heading of Sc. cremoris I have given several hints which will be of value in the production of starters for souring of cream. Particularly interesting are the cheese bacteria, and it was indeed the desire to study these further which led to the undertaking of the present work. Simultaneously with our bacteriological investigations therefore, we have carried out a great number of cheese-making experiments, in order to ascertain the effect produced in cheese by the bacteria found in cheese — and especially in Danish dairy cheese. The experiments were carried out with sterile apparatus and with so called iced milk from “The Copenhagen Milk Supply”. The milk was previously partly freed from germs, either by heating for half an hour at 70°, in the cheese vat itself, or by twentyfour hours’ treatment with 1.5 % hydrogen peroxide at 50° in an acid demijohn. The surplus of hydrogen peroxide was removed, immediately before adding the rennet, by means of Hepin, an extract of ox liver, containing catalase, placed on the market by the BEH- RING works, at Marburg. By this latter method, previously employed by GERDA TROILI- PETERSSON?), the power of the milk to coagulate with rennet is less impaired, but the milk 1) Centralblatt f. Bakteriologie. II. Abt. 1909, Bd. 24, p. 343. According to our investigations, nearly 1 “loo H»0s must be added to the milk in order constantly to prevent the development of yeast and bacteria. The hydrogen peroxide does not, however, impede the action of the proteolytic enzymes in the milk, and milk to which hydrogen péroxide has been added will therefore curdle in the course of 2—3 weeks, and then again dissolve. After six months, milk with 2°/oo HzO: was found to contain 79 °lo SN and 10.5 "Io DN. 111 189 is not rendered more free from germs than by heating to 70°. The bacteria of putrefaction were subdued in the experimental cheeses inoculated with lactic acid bacteria: whereas in the control cheese not inoculated, they developed so abundantly that these latter often contained more SN than the former. We were therefore obliged to disregard the results of the chemical investigations, and be content with a purely practical estimate, from taste and structure alone. If an experimental dairy should ever be established in Denmark, we hope to be able to resume the experiments with milk containing fewer bacteria. For the rennet, we used Hansen’s tablets, these being practically free from germs, which is by no means the case with the usual rennet extracts. As our experimental cheeses constantly proved better than our control cheeses, we can only say that all the lactic acid bacteria thus tested produce a favourable effect upon cheese, if only by subduing the putrefaction bacteria and furthering the proteolytic action of the rennet. An exception, however, is found in Sc. liquefaciens, which, as FREUDEN- REICH has already shown, renders the cheese bitter and makes it run. In cheeses with pure cultures of this streptococcus, we found up to 80 % SN. Tc. liquefaciens also can make the cheeses rather soft, but can also give them, as we have noted before, a flavour resembling that of “Danish Swiss cheese“, or “Russian Steppe cheese“. The best effects are obtained with Sbm. casei, which is a cheese ripening bacterium par excellence, but also the casein- splitting strains of Sc. lactis and Sc. cremoris gave an extremely good cheese, so that we can willingly agree with BARTHEL!) that a greater importance should be ascribed to such streptococci in the ripening of cheese than has hitherto been conceded. Even Sc. glyceri- naceus, Which does not break down casein, had a favourable effect on the ripening of the cheese. Sc. fecium, Sc. thermophilus, Sc. inulinaceus, Sbm. plantarum, betacocci and beta- bacteria do not as a rule affect the flavour. Vigorous strains of the two last-mentioned groups can occasion some formation of holes. The great importance attaching to certain thermobacteria, especially Tbm. helveticum, in the ripening of strong heated cheeses, is a point we need not here discuss further. In non-cooked cheeses, where cooling takes place rapidly, they cannot of course produce any affect. It will be seen then, that many different species of lactic acid bac- teria contribute to the ripening of cheese, but there is hardly any doubt that a cheese — as has been proved in the case of Emmenthal cheese — only obtains the proper character when certain definite species are pre- dominant. The bacteria to be used for the ripening of cheese must, like those used for souring of cream, be freshened up for a time in milk, and it will be necessary to make sure that the casein-splitting strains have not lost this power. 1) Meddelande Nr. 97 fran Centralanstalten för försöksväsendet pa Jordbruksomrädet. 1914. do No. 171. 1918. As regards the ripening process of cheese otherwise, I can refer to my Dairy Bacterio- logy 1916. Summary. 1. The present work is generally concerned with the true lactic acid bacteria, and more particularly with those which are of importance to the dairy industry. The strains described number 330 in all, most of these having been under observation for many years in order to ascertain the constancy of their qualities. 2. The true lactic acid bacteria form a great natural group of immotile, sporeless, Gram-positive cocci and rods, which in fermenting sugar form chiefly lactid acid. In a freshly isolated state however, some few species approach the pseudo lactic acid bacteria, the coli and aerogenes bacteria, by forming fairly considerable quantities of by-products. These latter consist principally of acetic acid and carbonic acid, at times also succinic acid; more rarely, mannite and hydrogen. That we nevertheless reckon such species among the lactic acid bacteria is due to the fact that they resemble them in all other respects, and gradually lose, partially or entirely, the power of forming by-products, thus ending by becoming true lactic acid bacteria. 3. As the sources of energy are utilised more completely where the quantity of by- products formed increases, the formation of these is a sign of vitality. An exception, however, is the formation of acetic acid, which is proportionately greatest under unfavour- able conditions, as for instance where air and acid abound, and at too high temperatures. 4. The lactic acid formed can be either dextro-rotary or levo-rotary. Where equal quantities of these two acids are formed, we obtain pure inactive lactic acid. Strains of bacteria which in milk form pure dextro-, or pure levo-lactic acid, will also in any nutritive broth likewise form respectively dextro- and lævo-lactic acid, whether the source of energy be alcohols, aldoses, ketoses, pentoses, hexoses or polysaccharides. Strains which in milk form equal quantities of dextro- and levo-lactic acid will as a rule also under other con- ditions maintain the equilibrium between the two acids, whereas strains which in milk form more of the one than of the other will under unfavourable conditions generally form only the one which they more easily produce. Even in milk, indeed, such bacteria may end by forming only the one acid, andit may happen without reducing their total pro- duction of acid. 5. As the modification of the lactic acid is entirely independent of the stereochemical structure of the sugars, and determined solely by the species of bacteria, we must suppose that dextro- and levo-lactic acid are formed each by their own particular enzyme. 113 191 6. In the pure lactic acid fermentation, the hexoses are simply split up into two molecules of lactic acid; the pentoses, on the other hand, are not, as might be expected, split up into one molecule of lactic acid and one molecule of acetic acid, but as a rule accor- ding to the formula: 6C,H,0; = 8C,H,O, +3C,H,0, 6 pentoses — 8 lactic acid + 3 acetic acid. which thus gives proportionately more lactic acid. 7. The enzymes which hydrolyse the disaccharides appear to be endo-enzymes, and we must therefore suppose that these sugars are taken in as such. There is consequently nothing to prevent the disaccharides from being better nutriment than the monosacchar- ides of which they are composed, and as regards the betacocci in particular, they can form slime from cane sugar, but neither from dextrose nor from levulose, just as various other lactic acid bacteria likewise form slime from lactose but neither from dextrose nor from galactose. In the latter case, however, the nitrogenous nourishment is also of importance, as the slime formation only occurs in milk. 8. Of the four hexoses: levulose, glucose, mannose and galactose, the last is as a rule that which the lactic acid bacteria find most difficulty in fermenting. Some few species however are altogether incapable of fermenting mannose, and will then often not attack salicin either. In the case of some few species, the power of fermenting mannose and that of fermenting arabinose are inversely proportional. 9. Only saccharose-fermenting bacteria are able to ferment raffinose. Bacteria which ferment starch can also ferment glycogen, and vice versa. The same enzymes therefore, are required to affect vegetable and animal starch. 10. The true lactic acid bacteria never ferment gum arabic, erythrite and adonite, and only rarely dulcite and inosite. The fermentation of the two last is in any case only slight. 11. In studying a bacteria, it is not sufficient merely to investigate which sugars it ferments at all, but the quantity of acid formed must be estimated accurately in order to determine in what order the different sugars are preferred. The order of preference of the sources of energy is our most important means of identification but may however, like any other character, lead to serious errors if employed alone. 12. Only few lactic acid bacteria can thrive altogether without sugar, but all of them grow as long as there is only a trace of sugar present. In the case of the streptococci, and the tetracocci, the sugar optimum lies between %—2 %. An exception, however, is formed ‘by Streptococcus cremoris, which, like the betacocci, prefers 5—10 % sugar. The sugar optimum of the rod forms lies between 2—5 %. A sugar concentration of 2 % is, however in all cases extremely favourable. 13. The lactic acid bacteria are as a rule but poorly supplied with proteolytic enzymes. Some strains have no effect at all worth mentioning either upon peptones or casein (the majority of betacocci, betabacteria and partly also of the species Sfreplobacterium plan- tarum); others affect peptones, but not casein (most of the streptococci, and all strains of the species Tetracoccus casei); others again both peptones and casein (most of the tetra- cocci and of the species Sc. lactis, Sc. cremoris, Sc. bovis and Sbm. casei). The active en- zymes are in all these cases endo-enzymes (probably erepsin), which act only in nearly D. K. D. Vidensk. Selsk. Skr., naturvidensk, og mathem, Afd, 8. Række, V. 2. 25 192 £ 3 114 neutral liquid. Only Sc. liquefaciens and some few tetracocci give off proteolytic enzymes in a living state. These liquefy gelatin and can act in a slightly acid liquid, but are most powerful, however, with neutral reaction. 14. The cocci which split up casein decompose it gradually through peptones to amino-acids; the casein-splitting rod forms, however, peel off the mono-amino-acids from the casein molecule without previous formation of peptone. From the peptones, the lactic acid bacteria appear to form a quantity of polypeptides, which are not precipitated by phosphotungstic acid. 15. The true lactic acid bacteria are incapable of breaking domi amino-acids, and they, therefore, in splitting up proteins, do not form more ammonia than is present in the protein molecule as such. 16. In accordance with this, the true lactic acid bacteria are, in canna to the pseudo lactic acid bacteria, unable to thrive with single amino-acids or ammonia salts as source of nitrogen, but demand a nitrogenous nourishment as complicated as do the an- imals, viz. genuine proteins or the entire complex of amino-acids therein contained. 17. As the lactic acid bacteria are as a rule not provided with ectoenzymes, the proteins must be given in a state of solution or in colloid form. Casein is particularly suitable in the finely divided form in which it occurs in milk. The digestion is, however, rendered easier as a rule by the addition of rennet. Among other genuine proteins, gluten and legumin (dissolved in sodium phosphate) can be used. Gelatin, on the other hand, is a very bad source of nitrogen. Many lactic acid bacteria will comparatively rapidly lose the power of utilising casein, just as on the other hand they can gradually accustom themselves to other sources of nitrogen. 18. Inactivised blood serum is, even with the addition of potassium phosphate, a poor source of nitrogen even for pathogenic streptococci. LIEBIG's extract of meat, and casein pepton, on the other hand, are excellent sources of nitrogen for all lactic acid bac- teria. The microbacteria, however, do not thrive so well with the meat extract. Yeast extract had a very specific action, being an extremely bad source of nitrogen for patho- genic streptococci, but by far the best nitrogen source for the thermobacteria and strepto- bacteria. Nevertheless, casein peptone is generally to be preferred, on account of its light colour. WITTE peptoneis a far poorer source of nitrogen than casein peptone, and as it gives abundant deposits with acid, it is ill-suited to solid substrates intended for culti- vation of strong acid formers. In agar stab cultures, for instance, the stab is rendered al- together invisible. 19. Many lactic acid bacteria do not grow at all with only % % Wille peptone, an- swering to 0.07 % N. All thrive the better, the greater the quantity of assimilable nitrogen at their disposal. The betacocci and betabacteria, however, are impeded by concentrations of 2 % N. In the case of casein peptone or yeast extract, the effect is not as a rule increased to any essential degree by using more than corresponds to 0.5 % N. For our nutritive substrates therefore, we always use the last mentioned sources of nitrogen at these con- centrations. 20. That the lactic acid bacteria thrive better with increasing concentration of ' nitrogen, is essentially due to the fact that the organic nitrogenous nourishment acts as a buffer, and the fermentation of sugar is therefore increased with increasing quantity of 115 193 nitrogen in the nutritive substrate. That Wille peptone is inferior as a source of nitrogen to casein peptone and this latter — in the case of the strongest acid-formers — again poorer than yeast extract, lies to some extent in the fact that the buffer action of these sources of nitrogen rises in the order given. 21. That the difference between the mentioned sources of nitrogen is not exclusively due to their different buffer action, however, is distinctly evident from the relation of the lactic acid bacteria to the sugars. Only a nitrogenous nourishment absolutely sufficient in all respects will enable the lactic acid bacteria to produce the many heterogeneous enzymes (invertase, maltase, lactase, inulinase, etc.) which are required to hydrolyse di- and polysaccharides. It is therefore necessary to know which are the best sources of nitro- gen for the different bacteria before we can judge which sugars they are able to ferment at all. 22. Sugar fermentation should not, however, be tested only With a good source of nitrogen, but also with a poorer one; this will give a more complete impression as to which sugars are preferred. 23. The lactic acid fermentation is impeded not only by the hydrogen ions, but also by the lactate ions. The better the buffer action of the nutritive substrate, the more will these latter make themselves felt. As, however, the lactate ions are not nearly so dangerous to the life of the lactic acid bacteria as the hydrogen ions, nutritive substrates with good buffer action should be used for preservation cultures. In agar with only 4, % dextrose and 44% nitrogen in the form of casein peptone, we have succeeded in preserving lactic acid bacteria unweakened for over three years, and this, be it noted, without re-inoculation. 24. As the aerogenes bacteria will, with a good nitrogenous nourishment, turn all the sugar into gas, and render the substrate alkaline, they are better preserved in sub- strates with less abundant nitrogenous food. The same applies to the fluorescent bac- teria, strong alkali formers, and the more aerobic micrococci. Many of these can be well preserved in water with 2 % soluble starch. 25. Like most other bacteria, the lactic acid bacteria are also able to grow with hardly perceptible quantities of inorganic salts. Potassium phosphate is the most important nutritive salt, and furthers the development of the lactic acid bacteria in increasing quanti- ties up to 2 %, or even more (this, however, with a single exception). The favourable effect of this salt, however, depends like that of the nutritive substrate, partly upon buffer action. This is very distinctly seen in the case of the aerogenes bacteria, which turn all the sugar into gas, with great quantities of potassium phosphate. » 26. There is considerable difference in the amounts of common salt which the different species of lactic acid bacteria can stand. As a rule, 2.5 % will produce no detrimental effect; certain species are slightly impeded, others advanced in their development thereby. With 5.5 %, all are impeded, and 10.5 % will in most cases stop the growth entirely. An exception, however, is formed by the tetracocci, which can as a rule stand 15.5 % commom salt. 27. Important characters in a bacterium are the minimal, optimal and maximal tem- peratures for its vital activity, as also the maximal temperature at which it can live at all. P 25° 194 116 28. The optimal temperature for the acid formation, as also for the proteolytic action often lies slightly below the optimal temperatures for growth. 29. The optimal temperature is not affected by the conditions of nourishment, or by the vitality of the bacteria. The minimal and maximal temperatures, on the other hand, are to a certain degree affected, and weakened strains therefore exhibit far steeper tem- perature curves than those whose vitality is unimpaired. 30. Some few strains of Sc. cremoris and of the betacocci can grow already at 3°, whereas Sc. bovis and the thermobacteria do not as a rule develop until past 20°. Sc. fecium and the thermobacteria thrive well at 47%—-50°; some thermobacteria, indeed, can even growat over 50°. Sc. glycerinaceus, Sc. liquefaciens, and Sc. thermophilus, as well as the coli and aerogenes bacteria, grow well at 45°. The majority of the true lactic acid bacteria on the other hand exhibit poor growth already at 3714—40°, and their optimal temperature lies at 30° or below this. The pathogenic streptococci and Bacterium bifidum thrive best at 35—-37°, and Sc. thermophilus, the thermobacteria and Betabacterium lon- gum at 40° or over. 31. In determining the maximal temperature for life (the death temperature), the number of cells which can stand the temperatures used should be noted, as it is of far greater practical importance to know how the majority of cells behave than what the few specially resistant individuals can stand. It is only an extremely insignificant number of cells in a bacteria species which can stand the so-called death temperature. 32. The death temperature lies, for the pathogenic forms, below 60°. The betacocci are killed, when not protected by slime, at 65°, and the most common lactic acid bacteria of milk, Sc. lactis and Sc. cremoris at 70°, while most of the other lactic acid bacteria can often stand 70—75°. The greatest power of resistance to heat is shown by Micro- bactérium lacticum, which does not always perish even at 85°. The duration of the heating in these experiments ‘was a quarter of an hour. 33. The true lactic acid bacteria — in contrast to most other bacteria — lack catalase entirely. An exception is formed by the more aerobic forms; the tetracocci and most of the microbacteria, but these two groups of bacteria cannot be reckoned entirely to the true lactic acid bacteria. . 34. The true lactic acid bacteria are, in contrast to the pseudo lactic acid bacteria, unable to reduce nitrate to nitrite. The tetracocci and microbacteria are here again excep- tions. It is, however, by no means all the hydrogen-peroxide-splitting lactic acid bacteria which reduce nitrate to nitrite, in closely related strains, one may reduce, and the other not. 35. The lactic acid bacteria have as a rule a great aversion to air. They form small colonies on plates, and only an extremely thin veil in streak cultures, and they grow evenly throughout the whole of the stab without any considerable surface growth. Some few thermobacteria even grow better deeper down, and Bacterium bifidum is obligatorily anaerobic. Exceptions are the microbacteria, which with good nitrogenous nourishment will as a rule exhibit some surface growth, and the tetracocci, which most frequently have a strong surface growth. 36. The true lactic acid bacteria are as a rule not chromogenic. Some few pathogenic streptococci can on casein peptone agar form a red colouring matter in the stab, and Strep- tococcus mastilidis forms an’orange colour in casein peptone broth withsoluble starch. Mi- "ve er oe a 7 ee eee ee in né th pe audit 117 195 crobacterium flavum develops an orange colour on the surface, and many tetracocci form yellow, brown, orange or even red colours on the surface. 37. The magnitude and appearance of the surface growth, as well as the formation of colouring matter, are among the most variable qualities in the bacteria. 38. When lactic acid bacteria are cultivated in milk, they are at their first stage of development surrounded by a more or less distinct capsule. This can, under certain con- ditions (in the case of the cocci only at lower temperatures) swell up and turn into slime. The power of forming slime in milk is, however, very variable, and even though it may occur more frequently in strains of one species than in those of another, it cannot be used as species character. We have often encountered this power in Sfreplococcus cre- moris and in the thermobacteria which form inactive lactic acid, and, in these cases, it was always in strains of particularly strong vitality. When, on the other hand, a strain of one of the other species has temporarily proved slimy, it has generally been defective in one respect or another. 39. The variations we have encountered in the lactic acid bacteria can as a rule be explained as a further development of already existing tendencies, or more frequently, as the result of weakness or degeneration. 40. The manner in which a bacterium is inclined to vary is often one of the most characteristic of all its qualities. 41. In a bacteria culture, most of the individuals are but slightly resistant, and as a rule weakened in one or another respect, and it is therefore necessary to inoculate abund- antly in order to make sure of transferring some of the individuals of unimpaired vitality, which mark the culture as a whole. 42. Consequently, next to unsuitable composition of the nutritive substrate, and too high preservation temperature, the chief cause of the frequent degeneration met with in laboratory cultures is the slight amount of the inoculating material generally used. 43. As shown in my paper “The Main Lines in the Natural Bacteria System”, out of the three morphological qualities of bacteria: shape of the cells, formation of spores, and arrangement of the flagella, the last is the one which should primarily be used. For we find that all bacteria which are able to live exclusively upon inorganic nourishment, and derive their energy chiefly through simple oxygen processes, have the flagella terminally set, whereas those requiring more complex organic food, and producing the more typical fer- mentation processes, have the flagella distributed throughout the whole of the cell. On this basis, therefore, I have divided the bacteria into two orders: Cephalotrichine and Peritrichinæ. As the true lactic acid bacteria have no flagella, we cannot determine their position in the system from their morphological qualities, but as they require as complic- ated food as do the animals, there is no doubt that their place is in the order of Peritri- chine. 44. By taking the arrangement of the flagella together with the biological characters as the first principle of division, we obtain within the same family of bacteria both sphe- rical and rod forms, as well as screw forms, and the old generic names, which only express the shape of the cells, are therefore no longer sufficient, but require to be supplemented by various prefixes. If, in any exceptional instance, we use the old terms, then it will at 196 118 any rate be necessary to restrict their meaning .By streptococci, for instance, we understand cocci which divide in one direction and form dextro-lactic acid. Cocci which divide chiefly in one direction, and form lævo-lactic acid, we have called betacocci,from their being most frequently found on beets. In the older bacteriology the generic name Diplococcus for streptococci, which fall away at once after division has already been relinquished; similarly also, the generic name Micrococcus should be discarded for sarcinæ without cohe- rence. The cocci which divide in several directions, and form lactic acid, we have given the generic name Telracoccus. Micrococci and-sarcine which do not form lactic acid lie, of course, outside the scope of the present work. 45. On the basis of the present investigations, we have succeeded in dividing the true lactic acid bacteria into 5 genera and 22 species. To these should further be added Bacterium bifidum, which from the true ramification of the cells occupies an exceptional position. Our system of division will be seen from the table on p. 106—107. Here is also put down the group B (with 2 genera and 8 species) of related bacteria, which cannot, however, be reckoned among the true lactic acid bacteria; they only mark the boundary on the one side, as the coli and aerogenes bacteria mark the boundary on the other side. 46. Simultaneously ‘with these investigations, we have made experiments with a view to utilisation of the isolated species. It was found that by inoculating of cabbage with Be. arabinosaceus and Sbm. plantarum, an excellent sour cabbage was obtained, and it was also found that most important bacteria in the ripening of the not strongly heated cheeses are Sbm. casei together with Sc. lactis and Sc. cremoris. In the case of some few varieties of cheese, Te. liquefaciens is doubtless also of importance, It has been shown in previous works that Tbm. helvelicum ranks first in importance for the cooked hard cheeses. INDEX Page Bacterium acidi propionici............... 49, 105 — ÉCHOS tc ca Sea ts 20, 29 — RO EEE 105 — CO A RE oie ais rn dee < 20, 42 [LEDE IP RON TOR RE 97 LEE ORNE SERRE TE 100 — CAC ICT 2 N 2 se 99 — COOL ea... 100 LES TER doo) TOME ON EN a 68 ISELQCOCENS ATADINOSACEUS: 5)... «2.0 0. ur... 74 = NTRS a. Seats tegen nee RE 74 CAS oe FE a ace ee 48 TORRES A TER ee nn Lure 36 CERES TERE ES een Seine oc 10 [RUNDE DC sd Sie nn cn DUO ne donc aie de à 108 REROBITeNntIRCALION una fee en cess 106 Methodessatsceultivation "2. Mosse eae o — IR VESEI PAROI men ice 9 — ESEBTESERVRTIONRAE ne on oc e 5 LUBERON =. cs a castes ee 101 Micrahacteriume flapum 02... 2. eek ole ss 103 — SIGE SEE SE 4. 101 — KEE TCT” ar a Zr ei 104 — HICSENTELICHID So... ee 103 WIE REG NEEL ee PT See re 73 Morphological’ features. u... eee 0... de 9 RAC He ARBOR PELIMCMUS ER. Lune acess 110 PE TEIN AW EP ES TELE NS 47 Remaining pathogenic streptococci........... 66 — saprophytic streptococci .......... 65 RÉ RER ce 8 Er MIRE 82 SGNTCESHONMCATUOW : =. 200. kedede. Va — - — concentration ....... 16 — - — manner of utilising........ 16 Page MOUTCES I OMTIÉETOSONE T2 occ ate Tr Tan 23 = - =< . concentration .. 2... 28 — - — manner of utilising...... 32 SIMEDEGDACLELIA: PA Le ten ase Soa eco 88 MINEPLODACLETINMACASEL:, ; ae Ys Sonics cs a ee 88 - plantarıim Bop wis ee 96 SELCDEQCUCC HR ian ee Aue seinen ee = ol SENERE GEDE BOBS re. or as ie Sere on oi 59 — CTEINOTISE hi A SEERNE SOLER 54 = POPE ie Meh BOP OE Mey AON vee 61 — QL CELINACCUS “tee a oe ae 62 _ ENLACES ay, are ee 60 — TASER RER RE Ce 52 — lide fans SS SEERE eile ee 64 — Mos bide ym, Te SEK NE SETS REE 57 — DU OG EN EST N ANE aoe 67 — thermoplalusen er SKER SS SØE 58 Summary ..... SE ere NA EN 112 SyStematisn mare cps ee een 106 aurocholate of sodium". 8.2.2.0. 47 Memperatune, «death! tie charset ok eee 43 = Maximal RE bres eater 39 — nent PS AE Cun NR ST 2 39 _ Optimal? yg. re. en en 39 NE TAC DC CE EN ee N EEE 76 Df CLRGCOCCUSs CASCIO 3 Ka ee. 80 — Pique fa clenis ERP RE RS 80 — ITU CODEN MALUS eis ere SER ANE Des 81 TRE RDPACIET IAE ce Er chers slats 82 Thermobacterium bulgaricum ............... 86 A D OS M RE CR OO ES 86 — helveticunl 2»... cn a Samen 86 — UG UT et rnc etre eto 86 — TAGEIST NE RE cy, ry ae 86 Varia bilitve tet rs ME I ani Boe Series de 49 CTIC ACID BACTERIA PEATES INDEX Plan Plan Bacterium acidi propionici........ LI Streptococcus fecium............. XV—XVI « Biplane. ESF 3 LI — glycerinaceus........ XVI—XVIII } Betabacterium breve. ...... a ae GA XLVI—XLVIII = inulinaceus........ . XIV À ; « caucasicum ........ XLVI — al tn se REC LE CINE I—IV « longnnie oe SEE =A" XLVII — liquefaciens ......... XVIII Betacoccus arabinosaceus ......... XXI—XXII — mastilädis.. 2. . 2:.. «- X 22 eMMGDIS\ 2.7. ee Cr XXIII—XXIV — BUDT ENES REN. XX Microbacterium flavum ........... L — thermophilus........ X—XII « NT er, Fr REESE RENEE XLIX Tetracoccus casei (Nr. 5) .......... XXVI « liquefaciens ....... L — liquefaciens (Nr. 9—11). XXVI « mesentericum...... L — mycodermatus (Nr. 31). XXVII Remaining pathogenic streptococci. XX © Thermobacterium bulgaricum ..... XXXIII Remaining saprophytic streptococci XIX — CEIBRIER nr XXVIII Streptobacterium casei ............ XXXV—XL | = helveticum ...... XXXI — plantarum...... XLI—XLV — SHOT EG ae XXXII Streptococcus bovis ............... XIJI—X1V — lactis ara) 2.222. oe VILI— SEX — BE ee TV —IX — må . Sens Sie sD Ve All the microphotographs are magnified 1000 times. The preparations are, where not otherwise stated, made with indian ink according to the method of Burri. W = Witte Peptone, C — Casein Peptone, Y — Yeast Extract. AG = Alkaline sugar-free C-Gelatin. SG == Neutral Dextrose C-Gelatin. Agar — Neutral Dextrose C-Agar. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla-Jensen). BIT, Streplococcus lactis. No.4 Agar Streak, 1 Day, 30°. No. 4 Agar Streak, 1 Day, 37,50. Nr. 4 Agar Streak, 10 Days, 100. No. 4 S.G.-Plate, 4 Days, 20°. Autotypi. Pacht & Crone’s Eitflg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PLAIT Streptococcus lactis. No. 7, C-Bouillon, 3 Days, 10°. Nr. 7, Agar Streak, 10 Days, 109, se" | = ” vs | wt = * + Re FY A; 5 à ~ SER | > „on n0ssse “ee PL No. 7, Agar Streak, 3 Days, 37,50. No. 7, A. G.-Plate, 5 Days, 20°. No. 17, C-Bouillon, 1 Day, 30°. Pacht & Crone's Eftiig. Autotypi. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla-Jensen). PI. III. Streplococcus lactis. i No. 13, Milk, 1 Day, 30?, Gram stained No. 13, C-Bouillon, 2 Days, 30°. No. 13, Agar Streak, 6 Days, 30°. À No. 13, Agar Stab, 6 Days, 45°. No. 13, AG-Stab, 6 Days, 20°. Autotypi. Pacht & Crone’s Eitilg. IV. Pl. (Orla-Jensen). D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 Streplococcus cremoris. DA at se tn, 1, C-Bouillon, 3 Days, 30°. No. Streptococcus lactis. Gram stained. 14, Milk, 2 Days, 30°, No. 309, 1, Agar Streak, 1 Day, No. X (7 a ° Ar a 14, Agar Streak, 1 Day, 30°. No. 10", 10 Days, 1, Agar Streak, No. 14, Agar Streak, 10 Days, 10°. No. o 5 's Eftilg. Crone Pacht & Autotypi. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PE V. Streptococcus cremoris. | : © de i i ws id x 5 | | ; . e CA i . É | g | Cd 7 > No. 5, C-Bouillon, 1 Day, 30°. -No. 5, Agar Streak, 1 Day, 309, No. 5, Agar Stab, 2 Days, 309, No. 5, SG-Plate, 8 Days, 209, Autotypi. Pacht & Crone's Eitilg, ’ + D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen) Plo Vi, penne POR - 3 . . ; gf À ; z f ” : ’ =" u Erz *, se Done 4 ir - py” M . . LA N / : = run net 1 No. 18, Milk, 1 Day, 30°, Gram stained "No. 18, Agar Streak, 19 Days, 10°. No. 18, SG- Plate, 4 Days, 20°. Autotypi. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). PL VII: Streptococcus cremoris. No. 19, Milk, 12 Day, 20°, Fuchsin. + >. } es a PL à AE « i - Ca à u 5 “ pe | à as fe Nr # AS) ‘ ‘ "Pu, bi we té , x cr x PR. « En No. 19, Milk, 3 Days, 20°, Fuchsin. No. 19, Milk, 5 Days, 20°, Gram stained. No. 19, C-Bouillon, 1 Day, 30°. No. 19, Agar Streak, 30 Days, 50. | Autotypi. Pacht & Crone's Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V 2. (Orla-Jensen), PI. VIII. Streptococcus cremoris. oes 1 A, a if 2 ET ~ i # . » Ca oe - + er L sem 7 Vin or . a LA s > + L . 5 Be ina Les . No. 20, Milk, 1 Day, 30°, Gram stained. uf TT. No. 20, C-Bouillon, 2 Days, 30°. No. 20, Agar Streak, 27 Days, 50. No. 20, Agar Streak, 1 Day, 30°. No. 20, Agar Streak, 1 Day, 30°, Fuchsin. Autotypi. Pacht & Crone's Eltñg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PIX Streplococcus cremoris. No. 20, SG-Stab, 6 Days, 20°. No. 20, AG -Stab, 6 Days, 20°. No. 2, C-Bouillon, 2 Days, 30 0. No. 2, Agar Streak, 1 Day, 30°. No. 2, Agar Streak, 2 Days, 37,5°. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). PI. X. Streptococcus maslilidis. No. 2, SG- Stab, 8 Days, 20°. No. 2, AG -Stab, 8 Days, 20°. No. 3, Milk, 2 Days, 300, Gram stained. Streptococcus thermophilus. No. 5, Agar Streak, 2 Days, 30°, No. 5, Agar Streak, 2 Days, 45°. | Autotypi. Pacht & Crone's Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). Pl. XI. Streptococcus thermophilus. No. 3, Agar Streak, 10 Days, 30°, No. 3, Agar Streak, 10 Days, 45°. No. 3, C-Bouillon, 4 Days, 45°. . No. 4, C-Bouillon, 4 Days, 45°. “a No. 4, Agar Streak, 4 Days, 45°, Fuchsin. No. 4, Agar Streak, 4 Days, 45°. eS Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2. (Orla-Jensen). PI. XII. Streptococcus thermophilus. ee ss . OR ? . i ” sø - ” * D 32 ac” ” x = ‘ DR i 2 v - >= os LA x . 2 Ye = ES 2 . . > 29 = “we a - er vs : nes > 5 Le z > a° & . ” > ne tow, EE een No. 2, Milk, 1 Day, 30%, Gram stained. No. 2, Milk, 1 Day, 45°, Gram stained. "meeen ener ae Tym Finn, - . >. >. > pt. L 3 - > monte, te No. 2, C-Bouillon, 1 Day, 30°. No. 2, Agar Streak, 1 Day, 30°. No. 2, SG-Plate, 31 Days, 20°. No. 2, Agar Streak, 1 Day, 45°. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. Autotypi. Streptococcus bovis. ~ * “see SE No. 1, C-Bouillon, 1 Day, 30°. W i x Lp Ÿ Fi A. Che ee a No. 5, Milk, 2 Days, 30°, Methylene Blue. (Orla-Jensen). PI. XIII. . No. 3, Agar Stab, 2 Days, 30°. No. 5, Agar Streak, 2 Days, 30°. Pacht & Crone's Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid., 8. Række, V 2. (Orla-Jensen). Pl. XIV. oo — No. 5, Milk, 1 Day, 30°, Gram stained. - No. 5, C-Bouillon, 2 Days, 30°. No. 5, Agar Streak, 1 Day, 30°. CVS » 8 . + À = : a, ? i s as: St an er - . 7 % 2 s = TER | 3 Streptococcus bovis and Micrococci in Cowdung, Gram stained. Streplococcus inulinaceus. Autotypi, Pacht & Crone’s Eftilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aïd., 8. Rekke, V. 2 (Orla-Jensen). PI. XV. Streptococcus faecium. “es % + - 5 - . Ea * : PTE - = ei > . 2 ee‘ F + = 4 - 3 0 = 7 = ré - i nes LA £ = a er q fm, ” A Rå Pr > £ FT et, RES = . = + ce: Pern et ae i D 2 =, es te . : - i FE u No. 2, Agar Streak, 1 Day, 30°. No. 3, Agar Streak, 1 Day, 30°. No..8, Milk, 2 Days, 30°, Gram stained No. 8, Agar Streak, 1 Day, 450. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). Pl. XVI. Streptococcus faecium. Streptococcus glycerinaceus. No. 1, C-Bouillon, 1 Day, 30°, No. 1, Agar Streak, 1 Day, 30°. A pn e er z a ’ j AE © i + ats Dy d Sey. ‘ ae ve an": 4 5 i eee LÆSES D. HU én ney : Be Le 5 i 4 .. - er 2 a es > år 7 ‘ > 1% $ . « - . s 2 ' 3 4 i No. 12, C-Bouillon, 1 Day, 30°. No. 12, C-Bouillon, 7 Days, 45°. - F i 2 . . \ . ik, - ce {= Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk, Skr., naturv. og math, Afd., 8. Rekke, V 2. (Orla-Jensen). PI. XVII. Streptococcus glycerinaceus. No. 5, C-Bouillon, 1 Day, 300. No. 5, Agar Streak, 1 Day, 30°. ns en À Autotypi. Pacht & Crone’s Eitilg. D kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PI. XVIII. Streplococcus glycerinaceus. No. 6, Milk, 2 Days, 30°, Gram stained. No. 6, W-Bouillon, 2 Days, 309, Streptococcus liquefacieus. No. 3, Agar Streak, 4 Days, 450. No. 5, C-Bouillon, 1 Day, 300. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). Pl. XIX. Remaining Saprophytic Streptococci. No. 1. Agar Streak, 1 Day, 30°. No. 7, Milk, 1 Day, 30°, Gram stained. No. 7, C-Bouillon, 1 Day, 30°. No. 7, Agar Streak, 7 Days, 10°. 00 ——— Autotypi. Pacht & Crone's Eitile. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla-Jensen). PI. XX. Remaining Pathogenic Streptococci. No. 6, C-Bouillon, 2 Days, 30°. No, 6, Agar Streak, 1 Day, 30°. No. 10, C-Bouillon, 2 Days, 30°. No. 10, Agar Streak, 1 Day, 30°. Autotypi. Pacht & Crone’s Eftilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid., 8. Rekke, V. 2 (Orla-Jensen). PI XXI Belacoccus arabinosaceus. No. 4, Agar Streak, 3 Days, 30°. j | No. 9, C-Bouillon, 2 Days, 30”. . No. 9, Agar Streak, 1 Day, 30‘ . 0 LEE i je) In > | ‘ 2 + N =~ - % s ® ' SS = or m Sy "> = . ' N | Ss ee res | Se = -- an / LA > 7 £ | Te. IN Sas u ehe 2 4 “ Lon! oF = ETES Du . N} | “1 NF her ae | A N Au y Pia: & s - À z À m oO ee es EN 4 - = SE x ; we 4 > Ne » , a : an N VÆR No. 12, C-Bouillon, 3 Days, 30°. No. 12, Beet Gelatin Stab, 1 Day, 20°, Fuchsin. | Autotypi. Pacht & Crone's Eitilg. D. kgl. D. Vidensk. Selsk, Skr., naturv. og math. Afd., 8. Rekke, V 2. (Orla-Jensen). PI. XXII. Belacoccus arabinosaceus. # r 4 writ cess PAY aS se = : | KI Pe FE 4 S Ye) en ke 2 åg; Vg AN » ae, ; V ; vv VE OA REDEN ROUTE ‘14; , ear, ann u ng \ LT Er 2. a as 7 aria 4 X x i me . “4 mp PT : se, - . ; Meee ee. 5? ” ” ur F4 No. 7, Agar Streak, 1 Day, 30 0. No. 7, Agar Stab, 2 Days, 30°. No. 11, Milk, 3 Days, 30°, Gram stained. No. 11, C-Bouillon, 3 Days, 30°. ‘J WIN ERE y % 1 > % A ; a > _ . a in \ À P} egg N . 7 X 4 + \ 2 “ee, ote À ” Te 2 ee Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math, Afd., 8. Række, V. 2 (Orla-Jensen). PI. XXIII. Betacoccus bovts. No. 33, C-Bouillon, 3 Days, 30°. No. 33, Agar Streak, 1 Day, 30°. No. 35, Agar Streak, 1 Day, 30°. No. 35, Agar Streak, 10 Days, 35°, Fuchsin. No. 36, Agar Stab, 2 Days, 30”. No. 42, C-Bouillon, 1 Day, 30°, Gram stained. Autotypi. Pacht & Crone’s Eitilg, D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla Jensen). Pl. XXIV. Belacoccus bovis. No. 44, AG-Plate, 8 Days, 20°. No. 46, C-Bouillon, 2 Days, 45° a © oes Lits æg" ‘on No. 46, C-Bouillon, 2 Days, 30 „Aller broat“, Saccharose Bouillon, 1 Day, Antimony Mordant and Fuchsin. From Zettnow. — ee nn. Autotypi. Pacht & Crone's_Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid., 8. Række, V. 2 (Orla Jensen). PI. XXV. Saccharose Gelatin Stab Culturs of Betacocct. 25 Days, 20°, Natural Size. Ss E 3 i = = 3 = o = o a [ee] 4 ZB —— en. 2 No. 35. No. 4. No. 5. No. 12. E = o Q (9) z 5 vo jo) © ov faa) C-Gelatin. Nos. 6 and 11 liquefy the Beet- and C-Gelatin with Saccharose. Nos. 11 and 12 prefer the Beet Gelatin. me u Dun 2 mg —m— Autotypi. Pacht & Crone’s Eitilg. D, kel. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen), Telracoccı. No. 11. Agar Plate, 5 Days, 30°. No. 11, C-Bouillon, 2 Days, 300. PI. Autotypi. Pacht & Crone’s Eftilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). PI. XXVII. Tetracocci. > Sa - . Ne aoe ? = % we 3 ee +) |. Fi ? . À é N : 2 > No. 29, Agar Streak, 1 Day, 30°. No. 31, Agar Stab, 28 Days, 20°. Autotypi. Pacht & Crone’s Eftilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla-Jensen). PI. XXVIII. Thermobacterium cereale. | No. 5, Agar Stab, 2 Days, 40°, Methylene Blue. No. 5, Y-Malt Agar Stab, 2 Days, 40°, Methylene Blue. Thermobacterium lactis. No. 7, Milk, 4 Days, 40°, Gram stained. No. 7, Agar Streak, 1 Day, 40°, Methylene Blue. Autotypi. Pacht & Crone’s Eftilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math, Afd., 8. Række, V 2. (Orla-Jensen).. Pl. XXIX. Thermobacterium lactis. No. 6, Agar Stab, 2 Days, 40°. No. 10, Milk, 2 Days, 40°, Gram stained. No. 10, Agar Stab, 1 Day, 40°. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). Pl. XXX. Thermobacterium lactis. Po. m Be Ø i x Me a 4 \ ” sø 4 0 » ©, € \ SS No. 9. Agar Stab, 1 Day, 40°, Methylene Blue. No. 11, Milk, 2 Days, 40°, Methylene Blue. No. 8, Milk, 3 Days, 30°, Methylene Blue . No. 8, Agar Streak, 1 Day, 40°. No. 8, Y-Agar Plate, 3 Days, 40°, Fuchsin. SS Autotypi. Pacht & Crone’s Eitilg. res + D. kgl. D. Vidensk. Selsk, Skr , naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). Pl. XXXI. ET ——…—"—…—"…"…"…"—"——. ——_— Thermobaclerium helveticum. No. 12, Milk, 1 Day, 40°, Gram stained. No. 12, W-Whey, 2 Days, 400. - No. 12, Agar Stab, 4 Days, 40°. No. 12, Agar Tube, 3 Days, 40°. a nn nn mn mn nn Autotypi. Pacht & Crone’s Eitflg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla Jensen). Pl. XXXII. Thermobacterium Jugurt. Nr. 13, Milk, 2 Days, 30°, Methylene Blue. Nr. 13, Milk, 2 Days, 30°, Methylene Blue. Preparation in Water. Preparation in Canada Balsam. — TL———————— ——— Autotypi. Pacht & Crone's Eitilg. eed D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid., 8. Rekke, V. 2 (Orla-Jensen). Pl. XXXIII. No. 14, Milk sterilised, 1 Day, 40°, Methylene Blue. No. 14, Milk, pasteurised at 80°, 1 Day, 40°, Methylene Blue. Thermobaclerium bulgaricum. No. 14, Agar Tube, 4 Days, 40°. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V. 2 (Orla-Jensen). mn nn nn Thermobacterium ?. Z © 5. Milk, 1 Day, 30°, Fuchsin No. 15. Milk, 1 Day, 30°, Gram stained. Nr. 15. i RE Ye { a mms | z à ~ 0 Nr. 15. Milk, 5 Days, 30°, Methylene Blue. Autotypi, Pl. XXXIV. Milk, 5 Days. 30°, Gram stained. Pacht & Crone’s Eitilg, XXXV. PI. (Orla-Jensen). og math. Aid., naturv. sk. Selsk. Skr., D. kgl. D. Viden Streplobacterium casei. 30 0. 2 Days, -Bouillon, C No. 5, 30°. 1, C-Bouillon, 2 Days, No. Agar Tube, 4 Days, s, 30°. 2 Day -Bouillon, » C 2 No. n u u ‘ ( £ 4 aA ay 5 = =) £ ju 3 ” A, f—} Lay I AG-Plate, 8 Days, 4 +, No. No. 4, Agar Streak, 2 Days, 30°. Autotypi. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PI. XXXVI. Streptobacterium caset. No. 7, Agar Streak, 2 Days, 30°. No. 7, SG- Plate, 8 Days, 20°. Auber É en je “ No. 8, Milk, 4 Days, 30°, Gram stained. Nr. 8, C-Bouillon, 2 Days, 30°. No. 10, C-Bouillon, 2 Days, 30°. No. 10, Agar Streak, 4 Days, 30°. —__—— ee 171717111777 lt nn Autotypi. Pacht & Crone's Eftilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid., 8. Række, V 2. (Orla-Jensen). Pl. XXXVII. es Streplobaclertum caset. No. 9, AG- Plate, 10 Days, 20°. No. 15, Agar Streak, 2 Days, 30°. | No. 16, C-Bouillon, 3 Days, 30°. No. 16, SG- Plate, 11 Days, 20°. | Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk, Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). Pl. XXXVIII. Streplobacterium caset. ee No. 18, C-Bouillon, 2 Days, 30°. No. 18, AG-Stab, 8 Days, 20°, Methylene Blue. | FÅ No. 28, C-Bouillon, 1 Day, 300. No. 29, Agar Tube, 3 Days, 30°. No. 31, Agar Streak, 2 Days, 30°. Autotypi. Pacht & Crone’s Eitflg, D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid .8. Række, V 2. (Orla-Jensen). PI. XXXIX. Streptobaclerium casei. No. 32, C-Bouillon, 2 Days, 30°. No. 32, C-Bouillon, 3 Days, 450. No, 32, SG- Stab, 15 Days, 200. - No. 32, AG-Stab, 30 Days, 20°. Nr. 33, C-Bouillon, 1 Day, 30°. Nr. 33, AG-Stab, 30 Days, 20°. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PI EXT: Streplobacterium casei. No. 34, Ropy Variety, Milk, 2 Days, 30°, Fuchsin. No. 34, Ropy Variety, C-Bouillon, 2 Days, 30°. py y ) ) No. 34, Agar Streak, 19 Days, 30°. No. 34. Agar Streak, 1 Day, 45°, Fuchsin. — EEE Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aîd., 8. Række, V 2. (Orla-Jensen). PI. XLI. Streptobaclerium plantarum. E * & £ s € ° ’ . à Ü = ; « . Pr No. 1, C-Bouillon, 3 Days, 30°. ‘ ; x 2 vi = fo N je : a \ — N — Te’ \ CN 4 4 4 No. 20, Agar Streak, 3 Days, 30°. No. 20, Agar Tube, 2 Days, 30°. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). Pl. XLII. Streplobacterium plantarum. No. 2, C-Bouillon, 2 Days, 30°. No. 2, Agar Streak, 2 Days, 30°. ar le r ot 2 3 ( = / Oem: ; \/ No, 18, Milk, 6 Days, 30° Gram stained. No. 18, C-Bouillon, 1 Day, 30°, No. 18, Agar Streak, 1 Day, 300. Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aid., 8. Række, V 2. (Orla-Jensen). PE XLIII Streptobacterium plantarum. | - ag - b ~~. »" / No. 5, Milk, 5 Days, 30°, Gram stained. No. 5, C-Bouillon, 2 Days, 30°. RE? * ‘No. 5, Agar Tube, 3 Days, 30°. No. 21, Agar Streak, 2 Days, 37,5°. Autotypi. Pacht & Crone's Eitflg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). Pl. XLIV. Streptobacterium plantarum. No. 30, Agar Streak, 3 Days, 37,5". 1 N Sa . ‘ . 4 i Nad AN f > >" UW Mn MAGE L sag. ar vi ; wee ~ i - øn “My Se mt 2% : do ke rh. U ig End oe No. 42, C-Bouillon, 2 Days, 30°. (] ns N = No. 43, C-Bouillon, 2 Days, 30°. No. 43, Agar Streak, 2 Days, 30°. | eS nn nn mn See Autotypi. Pacht & Crone's Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). Pl. XLV. ————— Streplobacterium plantarum. No. 44, C-Bouillon, 2 Days, 30°. No. 44, Agar Tube, 3 Days, 30°. No. 44, AG-Stab, 10 Days, 20°. No. 8, Agar Streak, 2 Days, 30°. Autotypi. Pacht & Crone’s Eitflg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Aïd., 8. Række, V. 2 (Orla-Jensen). Pl. XLVI. Betabacterium caucasicum. Belabacterium breve. No. 2, Y-Agar Tube, 10 Days, 30°. No. 5, SG-Plate, 8 Days, 20°. BERN 4 Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PI. XLVII. Betabacterium breve. No. 8, Agar Streak, 4 Days, 30°. ‘ LA i ” Ms LA é No. 9, Agar Streak, 3 Days, 30°. No. 10, Agar Streak, 4 Days, 30,°. No. 10, Agar Streak, 5 Days, 30°. No. 10, SG-Plate, 10 Days, 20°. eee eee ee eee ———mær—m— Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). PI. XLVIII. Betabacterium breve. | No. 11, C-Bouillon, 2 Days, 30°. No. 11, Agar Streak, 3 Days, 300, Betabacterium longum. No. 22, C-Bouillon, 1 Day, 30°. No. 22, C-Bouillon, 2 Days, 30°. No. 30, Agar Streak, 2 Days, 40°. Nr. 33. Agar Streak, 1 Day, 40°. ES 0 ————— Autotypi. Pacht & Crone's Eftflg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Rekke, V 2. (Orla-Jensen). Pl. XLIX. Microbacterium lacticum. No. 1. Agar Streak, 4 Days, 30°. No. 1. Agar Plate, 4 Days, 30°, No. 4. Agar Plate, 4 Days, 30°. No. 5. Agar Plate, 4 Days, 30° ae ee Autotypi. Pacht & Crone’s Eitilg. D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V. 2 (Orla-Jensen). PI. L. Microbacterium mesentericum. Microbacterium flavum. Nr. 7, Agar Plate, 3 Days, 30°. No. 10, Agar Streak, 2 Days, 300, Microbacterium liquefaciens. Agar Plate, 3 Days, 30°. Nr. 7, AG-Plate, 10 Days, 20°. a N ad “ - Cd 4 . 3 = 5 LA À RY Fest … : ?a lern No. 7, Agar Streak, 1 Day, 30°, Gram stained. No. 9, Agar Plate, 3 Days, 30°. s . Ss D LÉ 3 ee , 2 = 7 | . N | ; + > J 4 . ‘ 4 . Autotypi. Pacht & Crone’s Eitilg, I D. kgl. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8. Række, V 2. (Orla-Jensen). = ER Bacterium bifidum. From Infant Il, Agar Stab, 7 Days, 37,5". From Infant III, Agar Stab, 7 Days, 37,5". Baclericum acidi proptonict. x, Agar Stab, 2 Days, 30”. b, Agar Stab, 2 Days, 39°. Autotypi. Pacht & Crone’s Eitflg. ZOANTHARIA FROM SENONE AND PALEOCENE DEPOSITS IN DENMARK AND SKAANE BY K. BRUNNICH NIELSEN WITH 4 PLATES D. KGL. DANSKE VIDENSK. SELSK. SKRIFTER, NATURV. OG MATHEMATISK AFD., 8. RÆKKE, V. 3. ++) K+ KØBENHAVN HOVEDKOMMISSIONÆR: ANDR. FRED. HØST & SØN, KGL. HOF-BOGHANDEL BIANCO LUNOS BOGTRYKKERI 1922 PREFACE I; in this paper, I have succeeded in presenting many new species of corals to the public, the credit is largely due to the fact that a foundation for my work has been laid by others. Both of our own museums have possessed for a long time a large material, classi- fied and determined. I owe warm thanks to Professor A. JENSEN and Inspector, Dr. phil. MORTENSEN of the Zoological Museum, and no less to Professor Boaaitp and Docent Ravn of the Mineralogical Museum for their permission to make use of this. My thanks are also due to Professor, Dr. phil. GRONWALL of the University of Lund for the great interest he has shown in my work and for the direct help he has given me. Without his valuable assistance it would have been impossible for me to account for several of our single corals. 26* I. Introduction. Although Zoantharia belong to those fossils which were among the first to attract the attention of collectors in Denmark, both on account of the frequency of their occurrence in our chalk and tertiary deposits, and because of their striking appearance, yet not until very recently has any attempt been made to classify and describe these animal remains. FORCHHAMMER!)—?) mentions in nearly all his fossil series the appearance of one or two species of Turbinolia from Katholm, Møen, Faxe and the Cerithium chalk from Stevns Cliff, but he makes no attempt to determine these fossils more exactly. That it seemed desirable to classify them is proved by the fact that the Zoological Museum contains remains of corals, for the most part the stone kernel of single corals, partly classified and named as follows: Turbinolia faxoensis, — brevis, — ponderosa, — biseriata, — pusilla, — crassa, — ambigua. No small amount of work had been done with the corals as may be seen from a table, preserved in the archives of the Museum, in which is printed a list of Danish fossil corals. This was prepared as table III of a contemplated large work on Gaea danica. The two first tables, (tables I and IT) dealt with octocorals, and were published in my paper: Moltkia Isis, Stp. og andre Octocorallia (Vid. Selskabs Skrifter. Mindeskrift for Japetus Steenstrup XVIII). As it proved possible to find the originals of all the cuts shown in table III, this table is now being published in its original form as table III of this paper. !) FORCHHAMMER: Om de geognostiske Forhold i en Del af Sjælland og Nabooerne. Vid. Selskabs Skr. II. Del (Naturv. Skr.) 1825. *) FORCHHAMMER: Danmarks geognostiske Forhold. Indbydelsesskrift til Reformationsfesten. Koben- havn 1835. 3 203 No text to the tables was found, but I have endeavored to retain the old specific names, used, in so far as is possible, in their original meaning, and by publishing the work to give a visible proof of the interest in paleontology shown by an earlier generation. Some of the specific names were commonly known to the earlier paleontologists. FISCHER BENzoN)"), for instance, cites the following series from Faxe: Turbinolia sp. (faxeensis), also found at Stevns Cliff. Monomyces pusillus, Steenst. and Forch. — brevis, — — Astraea. Caryophyllia faxeensis Bech. (Calamophyllia fax. d’Orb. L. and Br. Jahrb. 1851, p. 102.) Cladocora. PuGGAARD?) mentions as the only coral from Moen Turbinolia excavata V. Hag. LUNDGREN?) gives the following list of corals in his fauna. Monomyces faxeensis M. U. H. — pusillus M. U. H. Oculina sp. Caryophyllia faxeensis M. U. H. A. v. KoENEN*) describes two species of corals found at Vestre Gasværk, Copenhagen: Trochocyathus calcitrapa, Sphenotrochus latus. and finally there is a far more detailed description of corals given by HENNIG5). Of Zoan- tharia he mentions: Dendrophyllia candelabrum Hng. Lobopsammia faxensis Bech. Both these species were known earlier, the first as Cladocora sp., the second as Ca- ryophyllia faxensis Beck. Among the single corals he mentions: Parasmilia Lindstromi Hng. 1) v. FISCHER BENZON: Ueber das relative Alter des Faxekalkes. Kiel, 1866, p. 19. 2) PuGGAARD: Moens Geologi, Copenhagen, 1851, p. 66. 3) LUNDGREN: List of fossil Fauna of Sweden. III. Mesosoic. Stockholm. 1888. 4) A. v. KoENEN: Ueber eine paleocene Fauna von Kopenhagen. Göttingen, 1885. 5) Faunan i Skaanes yngre Krita. III. Korallerne: Bihang til K. Svenska Vet. Akad. Handl. Bd. 24. Afd. IV, No. 8. 1894.. : 204 > 6 This he says is likewise found at Faxe. Parasmilia scanica Hng. Ceratotrochus supracretacea Hng. Of the species we find again quoted in Ravn’s?) list of fauna: Dendrophyllia candelabrum, Lobopsammia faxensis, Parasmilia Lindstromi, - excavata. If we make a tabular comparison of these facts about Danish Zoantharia, we find that until the present time only ten forms were known; of these only the six following have been described: Parasmilia excavata, v. Hag. Sphenotrochus latus v. Koenen, Trochocyathus calcitrapa v. Koenen, Parasmilia Lindstromi Hng., Lobopsammia faxensis Beck., Dendrophyllia candelabrum Hng. (See list, p:'7.) As may be seen, our knowledge of these Zoantharia is not very great. Only the species which are underscored have been described and pictured,while the others are merely known as names in a museum, and in a single case even this is doubtful. There is, for instance, nothing in the collection to support FISCHER BENzon’s Astraea sp., for there is no coral from Faxe of asteroid form. However, as HENNIG too, claims to have observed “two imprints of an asteroid form’, we are forced to believe that there is some ground for the assumption. It seems reasonable to suppose that specimens of Heliopora incrustans were meant, for when the colony is well-covered with calcite this species readily resembles an asteroid coral. All the other forms given in the list have been found to tally with the species in question. Il. Anatomical Conditions. Great are the difficulties encountered in determining fossil corals from Danish deposits. As a rule the coral is filled with mud or coated with calcite crystals both of which efface and hide the details by which it is possible to determine the family, genus and species. It often becomes necessary to cut through the coral, and only in more fortunate instances in which the mud may be removed do all the characteristics become visible. 1) MILTHERS: Kortbladet Faxe og Stevns. D. G. U. I Række, Kbhvn., 1908. 205 b998Jalouidns snyooridsy BOIUBOS BITWSELIET IUOIJSPUPT emuseægq Wyo vrrourqduy smjej Snyoorjouoyds edei}oyeo No wniqejepues eıpAydoapuac] sisuoxey er AydojdeHqH (4 sueysniour Bodo) snnsiquie snyoo1307e.107:) sNsseio SNYIOAJOZIUY Bjeriosiq erpmwusereg SIA9IQ UIIUISO[207) snyusnd snyoonrdsf sIsuoXey SNYVOAJOJLUS BSOI9PUOA BIFIUISO 007) UJEAVIXO BITILISO[90") OZ6T | WIØNSPUTT BI[IWISeIed f wnzqeropueo f naopja1001dns \ snys0u}0)010/) f valunas \ piplusning IW0.1]SPU1T DIJIUSDID umuqnjapuna \ \erpAydoupuoq) v27)fydospuaq f f stsuoxey sisuaxp{ | \erururesdogo’] | rununsdogo"] f BJBALIX9 ae BI[LUISBIe 8061 UARY | 6687 SIUUOH | | Bun] eyisnd | saoXumouowf sisuexvy | sooAuouo f 8881 u913punT J snD] \snyooujousyds S vdpapopo \snyynfooyooiz CS8T u9uU20Y UOA BI020pPE[) f stsuoxey \eqpAydodrey BOBIISY f sraoïq | \sooAurouop f J emsnd | \s»9Awouonf \ erpourqanL 998] uozuag] IOYosty ensiquiv S,2AUIOUO Bsselod sooAulouosy BYELIOSIA SI9AULOUON SIAoIq S99ÂUIOUOIN eipısnd s99AwouoN SISUIXLBF SodAULOUOJ eso1puod sP9AwouoN!. f ppansxa UJBAUIX9 S99ÂLOUON \ vnourqung IGST pievssnq £98T ‘III AL voluep Er) 206 8 In the main I follow ZiTrEL!) and VAUGHAN?) for the skeleton and anatomy of the corals. Only at a very few points does the material at hand differ from the customary. The determination of Columella, however, is worthy of note. This formation is classified by various scientists as Columella proper and pseudo-Columella. Columella proper is supposed to originate from a formation deposited in the kernel of the coral at the earliest stages of the latter’s development, and to have developed parallel with this in the direction of the calyx. Columella proper may be divided according to its appearance into: 1) Styliform C. formed by a massive column of lime; 2) lamella-like C. formed from a lamella-like lime plate on which the costae find support, and 3) fasciculata C. formed by a group of twisted or folia- ceous columns of lime. Pseudo-columella is the name given to those formed by the inner edges of the septa, which in some way or other unite in the center of the coral. This junction of the septal edges might lead to the formation of spongiose or tortile columella. VAUGHAN makes a distinction between lamella-like columella and columella proper. demonstrating that the first is formed by a single septum which sends a leaf-like extension towards the center on which the other septal edges find their support. The Danish material which shows such a septum, (Sphenotrochus granulatus, tab. 1, figs. 2425) wholly supports this view. Cutting shows that columella is a continuation of a septal edge, and as the lowest portion of the coral has no columella, it cannot be an inde- pendent formation branching out of the base of the coral. Styliform C. is found nowhere in the Danish material. Fasciculate C. is supposed to be found in certain of the species described: Caryo- phyllia danica, Brachycyathus parvus and others, but the material is so small that it was impossible to sacrifice any of it to cutting. | In all the other species among which columella is found, and where, according to the theories of the earlier authors, there should be a fasciculate C., i.e. an independent formation projecting from the base of the coral, a pseudo-columella only seems to exist. This is not located at the base of the coral, but is formed during the growth of the animal by the inner edges of the septa curving from side to side in many folds. The curves of the adjacent septa overlap and grow together forming a spongy mass with very regular folds and channels of communication. This columella increases in thickness towards the top. It can attain very considerable dimensions and projects from the bottom of the calyx. Forms like Coelosmilia and Smilotrochus, which authorities claim are entirely devoid of columella, have a growth on their lower portions, though not the very lowest, resembling a columella formation, indistinguishable from the calyx and only visible when the coral is cut. It appears then, that all the forms of Parasmilia, Ceratotrochus, and Epitrochus possess a spongy pseudo-columella, and not, as was determined earlier, a true fasciculate columella. 1) ZirreL: Grundzüge der Paleontologie I. Abtl. München u. Berlin. 1915. 2) VAUGHAN: Eocene and lower oligocene Coral Faunas of the U.S. Washington. D.C. 1900. 9 207 III. Stone Kernels. The determination of those remains of corals in which the skeleton proper is disinte- grated, leaving only hardened deposits from the inter-septal compartments and occasionally the impression of the outer surface of the coral, presents many difficulties. These remains, too, are the very ones which are found most frequently in our deposits from Danien and senone. They are very common in coral chalk from Faxe and by no means unusual in chalk from Saltholm and Aggersborggaard. Numbers of them have been found in Cerithium chalk from Stevns Cliff. These stone kernels from the white chalk are only preserved when they are deposited in flint which prevents their destruction. The few species of ramifying corals to be found at Faxe produce stone kernels that are easily recognized by their mode of ramification. It is only necessary then to substantiate this characteristic in orders to determine them. They may, however, be determined even without this. Stone kernels of Dendrophyllia reveal the porous nature of that coral, for all the pores in the theca and its lacunous border are filled with mud in such a way that the upper surface of the stone kernel is covered with an entire layer of a closely lined web of chalk which anastomoses frequently and greatly resembles the tissue of hydro-corals. Haplophyllia faxensis has elongated stone kernels with the fillings of the interseptal compartments sharply defined. Amfihelia Becki, which like all Oculinae has a tendency to form compact endothecal fillings immediately under the calyces, yields thin, elongated, slightly crooked stone kernels, resembling single corals. The stone kernels of single corals present many more difficulties. Stone kernels are formed when mud fills the interseptal compartments. The fillings of the calyces are, there- fore, very intimately connected with the stony mass surrounding the coral, and it becomes almost impossible to analyse these conditions more exactly. Cutting the axis of the coral crosswise gives some information about the inner parts of the septum and the upper parts of the pali and columella; on the other hand cutting destroys part of the stone kernel, so it is advisable to have a large material on hand before resorting to this measure. Cutting lengthwise gives information about the lateral surfaces of the septum, about the upper edge of the septum and about columella. Among the endothecal formations the dissepiments are recognized by the fact that they hinder the further development of the stone kernel in the interseptal compartments in question, preventing, indeed, the mud from penetrating into the said compartments. If there are many dissepiments present, the stone kernel is short in comparison with the length of the coral, which may be measured by the impression of the theca. If, however, mud succeeds in penetrating beneath the dissepiments, the stone kernel is divided into several sections which lie one over the other. Other endothecal formations are difficult to locate, for they disintegrate together with the skeleton proper, leaving no trace. A central formation in the very depth of the coral may occasionally be determined, as for instance incipient columella formation. D. K, D. Vidensk. Selsk, Skr., natury. og mathem. Afd., 8. Række, V. 3. 27 208 10 The same is true of the exothecal formations. These disappear and the impression left in the chalk is the only clue as to whether the epithecum was present or not. Costa, on the other hand, leave visible traces. The matter may be summed up thus: The impression determines the outer surface of the theca. = —— eos ede, = — epithecum. — form of the coral. The stone-kernel determines the number and proportionate size of the septa. — presence of dissepiments. — central formation in the depth of the coral. Cutting the stone kernel lengthwise determines the upper septal edge. — lateral surface of thesepta. — columella. Cutting the stone kernel crosswise determines the conditions of the calyx. — the pali. — columella. IV. Taxonomy. In this paper the arrangement adopted by ZırreL: Grundzüge der Paleontologie, I. Abth. 4 Udg. 1915, is followed. It must be noted that the determination often rests on a very loose foundation due to the insufficiency of the material. In a few cases certain characteristics are supposed to be present even though the material at hand has been unable to furnish the proof. Possibly future collections will modify the determinations made in regard to the position taken by the species here described. The chemical condition of the lime skeleton seems to be one of the important points to be observed in systematizing the specimens. In some cases this seems to consist of cal- careous spar, in others of arragonite, as Johnstrup has shown in his paper on the formation of Faxe chalk. One species which seems to have a calcareous spar skeleton is Epitrochus pusillus, found everywhere with its skeleton in a good state of preservation, whereas other corals from the same locality only appear as stone kernels. The skeleton of this species must therefore be formed of calcareous spar, while the species which have disappeared were of arragonite. 11 209 V. The species grouped according to localities. From white chalk at Stevns Cliff: Coelosmilia excavata, Parasmilia cylindrica. Møen: Coelosmilia excavata. Aalborg: Coelosmilia excavata, — ponderosa. From Cerithium chalk at Stevns Cliff: Trochocyathus hemisphaericus, - Parasmilia biseriata, — cincta, Coelosmilia excavata. Besides these single corals there have also been found some remains of ramifying corals. However, these remains are few and small, and do not yet justify closer determination. Of fossils of older Danien we know from Stevns Cliff: Epürochus vermiformis, Parasmilia parva. Kagstrup: Parasmilia parva. Vixö: Epitrochus vermiformis. Bulbjerg: Epitrochus vermiformis. Of fossils of younger Danien we know from Rejstrup: Parasmilia parva. 27* Frederiksholm: Ceratotrochus Saltholmensis. Saltholm: Flabellum calcitrapa; Ceratotrochus Saltholmensis, Coelosmilia brevis. Aggersborggaard?): Epürochus pusillus, Smilotrochus faxöensis. Bryozo chalk at Faxe: Ceratotrochus Saltholmensis. Epitrochus pusillus, Parasmilia danica. Coral chalk at Faxe: Haplophyllia faxensis, Dendrophyllia candelabrum, Sphenotrochus granulatus. Ceratotrochus ambiguus. Epitrochus pusillus. Brachycyathus parvus. Caryophyllia danica. Coelosmilia brevis, Parasmilia Lindstromi. Smilotrochus faxoensis, Rhizotrochus crassus, Amfihelia Becki. Coccolith chalk from Limhamn: Parasmilia scanica. 1) Chalk from Aggersborggaard which contains a fauna with countless typical ‘‘Faxe’’ fossils is doubt- less a coral chalk formed in the same way as the coral chalk at Faxe but at a different stage of development, for here not only the corals themselves but their stone kernels as well have desintegrated and vanished. The corals can then only be determined by their impressions made on the surfaces to which they were attached in the case of Ostreae, serpulae and others. In this manner Dendrophyllia candelabrum and some hydrocorals are found again. 13 211 Hydro-coral chalk from Limhamn: Haplophyllia faxensis. Dendrophyllia candelabrum; Ceratotrochus ambiguus, Ceratotrochus Milthersii. Epitrochus pusillus, Epitrochus supracretacea, Coelosmilia brevis, Parasmilia Lindstromi, Smilotrochus faxoensis, Amfihelia Beckii. Herfölge gravel chalk: Ceratotrochus Saltholmensis. Copenhagen harbour: Ceratotrochus Saltholmensis, Flabellum calcitrapa, Sphenotrochus latus. Ravnstrup?): Ceratotrochus Saltholmensis. “Vestre Gasverk’’, Copenhagen: Flabellum calcitrapa, Sphenotrochus latus. As these lists show, a sharp distinction exists between the senone and the Danien species. The distinction between older and younger Danien is quite clear, only one species being common to both, Parasmilia parva. Most of the forms are common to the three divisions, the chalk formation, the tertiary formation and the present time; two, Dendro- phyllia and Amfihelia are only known from the tertiary formation and the present time. 1) The chalk from Ravnstrup is a gravel chalk of the same species as the upper chalk at Herfölge. Its content of fossils resembles that of Herfölge at every point. 14 VI. Schematic list of the species and their localities. Senon ‘Zone with Danien a MAR ur Younger Danien Be ie: ee] | Im E al | wl ol BHI 2 = 15: HEE d 5 BER ERBE FRREEIREBEIFE GE | =| =| 2] 2| 31>) Si 3] 8] S| 55|3|28 3] E] Gl ol = SAS AIS 2S58ae mg als | 16 | | QU es | > 1. Haplophyllia faxensis, Bech.............. | ; | : | ERA | +). 2. Dendrophyllia candelabrum, Hennig ...... | le + | +| Ile 3. Sphenotrochus granulatus, n.sp........... | : | - ke | EM 4. Sphenotrochus latus, v. Koenen........... | | : ++ 5. Ceratotrochus ambiguus, Forchh. & Steenstr. Weel | Re +|.!. En OAI | 6. Ceralotrochus saltholmensis, n. Sp.......... is | ; Le |+y)+}+j+ EUR 2. Geratofrochus, «Milthers en SD nn El |. + ++ 8. Epitrochus vermijormis, n.Sp. er. see +|l.|+-H!. = 9. Epitrochus supracretacea, Hennig......... | us + le 10. Epitrochus pusillus, Forchh. & Steenstr.... ; +)+)+ +]. |). | : 11. Trochocyathus hemisphaericus n. sp. ...... se + 3 12) Brachycyathius PAUSE Sp Terre ee — 135 (Gargophgilia \danica, ns peer EIGENE || + | 14."Coelosmiliaexcavald ANSER TR cm sei rele ah +|+ | | 15. Coclosmilia ponderosa, Forchh. & Steenstr. |. | .|+1.]- | 5 5 i || 16. Coelosmilia brevis, Forchh. & Steenstr.....]. | sl hs | + + +]. || 17. Parasmilia biseriata, Forchh. & Steenstr. . ae |. | | 18. (Parasmiliaycinclasem ssp KEE tenet oe | ed KER FA LE | | 19: 3Parasmiliayparva NES ER ES TES er : | i Se ern == 20; -Parasmilia cylimadnicas maspıe ie EC CE +| 2 2 21. Parasmilia danıcan. Sn TE ere aie + |. 22. Parasmilia Lindstromi, Hennig .......... | == 23. Parasmilia scanica, Hennig.............. | | M: TE HR 24. Flabellum calcitrapa v. Koenen .......... ; | le TB (es, Go | + [a ++ 25. Smilotrochus faxensis, Forchh. & Steenstr. . - | e |. [+ +} le | > + 26. Rhizotrochus crassus, Forchh. & Steenstr.. . | | LS aN age 3 27. Arofihelia ‘Becki, n. spi. ase © eit ee) ee ee +]. 1. | SENTE 2111214/2]1|1/1|1]3/12,3|1]3|1]1]1/11}1/2|2 Caryophyllia faxensis Beck 1835 in Lyell: Cretaceous and tertiary strata of the Danish VII. Separate Species. 1. Haplophyllia faxensis, Beck. Table I. Figs. 1—7. Islands. Geol. Transact. Ser. 2. Vol. 5, p. 249. Fig. 4. 15 213 Calamophyllia Faxoensis Beck 1848. M. Alcide d’Orbigny: Prodrome de Paléontologie vol. Il, p. 295. Calamophyllia Faxoensis, 1851, d’Orbigny, Leonhardt. et Bronn: Neues Jarhb. f. Minera- logie Jahr. 1851, p. 100—102 (Synopsis of A. d’Orb. Ueber die fossile Reste des Terrain Danien oder T. pisolithique). Rhabdophyllia faxensis 1854. Milne, Edw. et Jules Haime: Recherches sur la structure et la classification des polypiers récents et fossiles. Annales des sciences nat. 3 ser. Zool. V. IX—XVI. Caryophyllia faxensis Beck 1866. (Calamoph. f. d’Orb.) Fischer Benzon: Ueber das relative Alter des Faxekalkes. p. 19. Caryophyllia faxensis Beck. 1888. Lundgreen: List of fossil fauna of Sweden. III Mesosoic. Lobopsammia faxensis Beck. 1899. Hennig: Faunan i Skaanes yngre Krita III. Korallerne. Bih. til K. svenska Vet. Akad. Handl. Bd. 24. Afd. IV. Nr. 8. Stockholm. S. 11. As the list of synonymia shows, this species was known in early times. This is not sur- prising for it belongs to those corals which give the coral chalk at Faxe its characteristic appearance. Its determination, however, proved difficult, but as may be seen, d’Orbigny had already noted it as a Calamophyllia. However, as it does not possess the exothecal extensions, (Collerette), which distinguish genus Calamophyllia it must be classified with the closely related genus Haplophyllia. HENN1G determines it as a Lobopsammia for he presupposes it to belong to the porous coral group; however, this conclusion rests on an error doubtless due to the insufficiency of his material. Only stone kernels and impressions were at his disposal and so it was difficult for him to distinguish this form from Dendrophyllia candelabrum. His sketch of a cross- section of a Lobopsammia branch (Plate I, fig. 16) is, as a matter of fact, a cross-section of a Dendrophyllia branch identical with fig. 10 of the same plate. Indeed the cross-section plainly shows the coral tube to be a non-porous coral entirely lacking the lacunous edge so characteristic of the Dendrophyllia. The coral forms bushy colonies of considerable ‘size, characterized by dichotomous self-dividing ramifications. Nowhere are the epithecal formations known as collerettes to be seen, but occasionally plate-like formations uniting some of the branches appear (table I, fig. 1). The branches in the same colony increase but slightly in circumference so that there is very little difference between the older and younger ones. A single individual is located at the extremity of each branch. The calyx is rather deep, broad and cup-shaped, for the septa in the upper part of the calyx project only slightly from the thickened sharply truncated edge of the calyx. The theca is quite thick without epitheca, usually smooth, but sometimes, especially at the extre- mities of the branches, slightly protruding granular costae are seen. In the upper part of the calyx the septa are only slightly distinguishable from the edge of the theca, further down they protrude from the walls of the calyx, while at the bottom of the deep calyx they intercept and coalesce with its inner edges. The number of septa varies from 30—40, being as a rule about 36. It is usual to find 214 | 16 3 complete cycles (24 septa) and a fourth, incomplete. The single septa greatly resemble each other so that it is impossible to judge from their appearance to which species they belong. As a rule only two kinds of septa are found in a calyx, a comparatively large one alter- nating with a comparatively small one. The lateral surfaces of the septa are almost smooth, though diagonal rows of slightly protruding granulations may be seen running inwards and downwards. The upper edge of the septum is uneven, finely dentate, further down the edge is wavy. Dissepiments often occur in the interseptal compartments, many appearing at the same height. The hollow beneath the calyces fills up quickly as a rule, but sometimes we find long pieces of the trunk with open interseptal compartments without dissepiments. There is no columella. In the deeper sections of the trunks however, we observe that the septa coalesce by the help of their slightly billowing edges, turning towards the center, without though forming an ostensible pseudo-columella. A cross-section reveals septa from one side of the theca forming an easy transition with septa from the other side (table I, fig. 3). New individuals are formed by splitting. The mother individual increases in breadth and number of septa; it then contracts in the middle and two equally large individuals are formed. The colony may have a considerable extent but the trunks are apparently very little thicker in the lower parts than in the upper parts. The colonies seem to prefer to spread out like a fan, the branches dividing on the same plane. This characteristic is common to this form and to the majority of the Octo- and hydro-corals of Faxe. Locality: Danien: coral chalk from Faxe (also known from Annetorp). 2. Dendrophyllia candelabrum, Hennig. Table I. Figs. 8—22. Cladocora. 1866. v. Fischer Benzon: Ueber das relative Alter des Faxekalkes. p. 19. Cladocora. 1867. Johnstrup: Om Faxekalken ved Annetorp i Skaane. Overs. over det kgl. danske Vid. Selskabs Forh. f. 1866. No. 6. p. 9. Dendrophyllia candelabrum. 1899. Hennig: Faunan i Skaanes yngre Krita. III. Korallerne. Bih. til K. Svenska Vet. Akad. Handl. Band. 24. Afd. IV. No. 8. Dendrophyllia candelabrum. 1908. Ravn’s Faunaliste i Milthers: Kortbladene Faxe og Stevns. This species, which is the most common coral species to be found at Faxe and forms the largest part of the coral chalk there, was for many years determined as a Cladocora, because its outward appearance presents many points of resemblance with a Cladocora species from English tertiary formation. Much credit therefore is due to HENNIG, who with extremely insufficient material at his command, (only stone kernels and impressions), was able to see that its internal construction consigned it to the porous coral group, thus determining it as a Dendrophyllia. The same determination had, in the meantime, already been made by STEENSTRUP and FoRCHHAMMER, as is proved by drawings of that species preserved in the 17 215 collection of the Mineralogical Museum in which the porous character of the skeleton is apparent, and on the wrapper of which is written: “Dendrophyllia.’’ However, this fact in no way diminishes the credit due to HENNIG. The coral forms woody, ramifying colonies. The older stems are slightly thicker than the younger branches, but the difference is not great. The branches are formed by gemmation from the outer side of the theca without any connection with a calyx. The new individual, the new branch, first grows a slight distance at right angles from the parent stem, then turns and continues its growth parallel to the parent stem. Branches may grow from every side of a stem (table I, figs. 19—20). If two branches collide during their growth, a frequent occurrence under intense furcation, they merge into one another, forming a complicated net-like growth (table I, fig. 12). The theca is quite thick and, on account of the pores which are found throughout, forms a spongy (lacunous) tissue making a kind of connection between the interseptal com- partments and the outside world. The exterior is at most smooth, as a rule, however, it is possible to distinguish slight costae which confine the porous parts of the tissue. Occasionally the costae are more plainly marked and divide the outer surface into facets (table I, fig. 10). The calyx is quite deep and capacious, the free septal edges projecting only in the hollow of the calyx and to a very limited extent. Not until the bottom of the calyx are the septa able to reach the central formations. The free septal edge does not extend beyond the thecal edge; it is practically unbroken and is not dentate. There are between 24 and 36 septa; 3 complete ordines and a fourth more or less in- complete. The septa of the first and second ordines reach the center and by a coalescence of their inner edges form a pseudo-columella of spongy consistency (table I; fig. 8). A cross- section reveals the way in which this is formed from connate septal edges and shows that it is not a true independent columella (table I, fig. 16). A lengthwise cut through the stems shows how the interseptal compartments close to the calyx are closed by dissepiments. No endothecal formations are seen. The stone kernels are readily distinguishable from the twisted fillings in the pores of the lacunous edges which are located within the inter- septal fillings. i Locality: Danien. Coral chalk from Faxe and Aggersborggaard. (Also known from Annetorp). 3. Sphenotrochus granulatus, n. sp. Table I. Figs. 23—25. This species is a single coral of very small dimensions. It is about 10 mm long with a diameter at the edge of the calyx of about 2—3 mm. It is almost cylindrical in form, for its lateral development soon ceases and the breadth remains constant during the further growth of the coral. The axis is bent so that the coral is either curved once or in an S-form. The coral spreads a little at the base, forming an adhesive disc. D. K. D Vidensk. Selsk. Skr., naturv. og mathem, Afd. 8. Række, V. 3. 28 216 18 The theca is quite thick; on the outer surfaces are slightly projecting costae. These are broad and dotted with several rows of small granulations. There is no epithecum. There are 24 septa in 3 ordines and 3 cycles. Six in the first ordo are larger than the others, but as far down the calyx as may be seen they do not form columella. Cutting crosswise reveals a lamella-like columella, formed by the elongation of a single septum. This proves the species to belong to the genus Sphenotrochus. The upper septal edge is smooth and projects slightly beyond the edge of the calyx. The lateral surfaces of the septum are practically smooth. Many dissepiments are seen in the interseptal com- partments and the upper part of the coral is readily distinguished from the lower. Locality: Danien: Coral chalk at Faxe. 4. Sphenotrochus latus, v. Koenen. Sphenotrochus latus 1885 v. Koenen: Ueber eine paleocäne Fauna von Kopenhagen. Göt- tingen. S: 106. Tab: VW: Figs12/a, bte: This species belongs in certain cases to those fossils which give the chalk its special character. Wherever it is found it occurs in great numbers. In small loose blocks from Ystad and Halsted I have seen it in enormous quantities. I have nothing further to add to v. Koenen’s description. Locality: Younger Danien. Copenhagen harbor; Vestre Gasværk. 5. Ceratotrochus ambiguus, Forchhammer and Steenstrup. Table III. Figs. 1b, 5 and 6. This species belongs to those commonly found in the coral chalk from Faxe. It appears as stone kernels, but there are a few remains of impressions which permit a description of its external appearance. The stone kernel is obovate-conical in form, gradually pointed toward the base of the coral. The axis is sometimes straight, but as a rule bent either once or in the form of an S. The coral may attain a length of 35 mm with a diameter at the edge of the calyx varying between 12—18 mm. The stone kernel shows traces of 50—60 septa in 6 ordines containing 1 incomplete and 7 complete cycles. The septa from the second and in part from the third ordo as well, are large and well-developed, whereas the interjacent septa are small. The interseptal com- partments of the stone kernels, then, are grouped in 4’s. The single groups are separated by deep furrows thus rendering the stone kernels easily recognizable. A calycinal cross-cut shows that the calyx is quite deep. A cut further down reveals loosely united inner septal edges. There was no true columella and there were no dissepiments, for the stone kernel extended without break to the base of the coral. The lateral surfaces of the septum were covered with rows of such small granulations as to appear practically smooth. 19 217 The impressions and the few remains still existing show that the theca was thin, smooth without costae, and without epithecum. As I was unable to find any special difference in the stone kernel shown in table III, fig. 1b, 5 and 6, I have grouped them all under one of the names. Locality: Danien. Coral chalk, Faxe, Limhamn. 6. Ceratotrochus Saltholmensis, n. sp. Table I. Figs. 26—31. This species is a very small one. It attains a length of only 12 mm, with a diameter at the calyx of 5 mm. It is regular in form, obovate-conical. There is no basic extension, no stem, but sometimes a small adhesive disc appears at the point; if this is not the case the point is smoothly rounded. The theca is quite thin, without epithecum. The costae in the primary cycles are rather conspicuous, form no continuous ridge, but are scattered in rows of small prickles. The calyx is shallow with a prominent, rather broad columella filling about one third of the diameter. There are between 32 and 36 septa in 1 incomplete and 3 complete cycles on 4 ordines of which the 2 first are considerably developed and share in the formation of the columella. The later series are but weakly developed. The septa do not protrude beyond the edge of the calyx, they soon bend their upper edge down into the depths of the calyx, where, their edges uniting in the center, they form a spongy pseudo-columella capable of considerable development (table I, fig. 29). The septal edges seem to be entire. The lateral surfaces of the septa are covered with granulations, often quite large, spinate, which greatly reduces the interseptal compartments. No dissepiments are found, but the interseptal compartments are unencumbered down to the lowest parts of the coral. The inner edges and the columella are much thicker on the lowest parts of the coral and some solid endothecal chalk deposit is visible, making these lowest parts of the coral almost massive. It is not unusual to find specimens in which the larger portions of the theca and septa are worn away and only the lowest part of the coral with the thickened end of the columella preserved (table I, fig. 30). Locality: Danien: The species is well known from younger Danien; not known from older Danien. Younger Danien: Frederiksholm, Saltholm, Bryozoa chalk from Faxe. Her- folge, Ravnstrup, Copenhagen harbour. 7. Ceratotrochus Milthersii, n. sp. Fig. 1—4. Among the most common fossils in the paleocene deposits near Randers is a coral which only appears as a stone kernel with no impressions. Various cuts showed the coral to have about 30 septa which are contiguous in the center of the coral and form small spongy columns. The interseptal compartments were free without dissepiments. At the lower end 28* 218 20 of the coral the septa were much thicker and the interseptal compartments thus partially closed. The sides of the septa show the usual curved rows of granulations but without larger thorns. None of the specimens showed any traces of theca. The coral was obovate-conical with a slightly bent axis and there was no sign of root extension. In certain respects this Fig. 4a and b. Fig. 1. Ceratotrochus Milthersi. Stone-kernel. Side view. * 1. Fig. 2. do. do. do. Cross-cut. “/1. Fig. 3. do. do. do. Tangential cut *ı. Fig. 4aandb do. do. do. Cross-cut and side view ‘/1 of the lowest end. coral suggests Ceratotrochus Saltholmensis, but may readily be distinguished from that coral by its far greater size and by the appearance of the sides of the septa. A few rare specimens of similar stone kernels may be found at Faxe and Limhamn. In loose blocks of the same age from Halsted near Nakskov are found two corals with their shell preserved. They correspond at every point to the stone kernels referred to here, and I do not therefore hesitate to group them with these. They further strengthen the sup- position that the coral was a Ceratotrochus. The under side shows traces of slightly protruding costae covered with countless granulations. These, however, are not arranged longitudinally as is the case with C. Saltholmensis. The two specimens show only the lower part of the coral, so it is impossible to describe the calyx. Locality: Upper Danien: Faxe, Limhamn. Green sand chalk near Randers. 8. Epitrochus vermiformis, n. sp. Table II. Figs. 1—4. This species is characterised by its epithecum, which not only entirely covers the out- side of the coral, but also any objects with which the coral comes into contact during its growth, — for instance bryozoa colonies (table II, fig. 1). 21 219 The coral forms a root extension of considerable size in the inside of which are seen the remains of the bryozoa colonies or similar growths on which the young coral sought support. The coral has the usual obovate-conical form, but in many places the growth is but slight so that the form becomes cylindrical. The axis is always bent, usually in several direc- tions, giving the coral a worm-like appearance. The theca is smooth, without corticles, but as the epithecum in its growth surrounds everything with which the coral comes in contact, the theca is sometimes rough with pro- tuberances of various shapes. (table II, fig. 1). The calyx is deep. There may be as many as 40 septa, 5 ordines in 4 cycles of which 12 reach the center while the others are but slightly developed. The septa do not extend beyond the edge of the calyx but soon disappear into the calyx. The septal edge is curved. The lateral surfaces of the septum are smooth. No rows of granulations are visible. The inner septal edges unite to form a pseudo-columella which attains considerable thickness towards the top; the upper part may be seen at the bottom of the calyx. There are no dissepiments or other endothecal formations. Locality: Danien: Older Danien: Stevns, Kagstrup, Bulbjerg, Vixo. 9. Epitrochus supracretacea, Hennig. Ceratotrochus supracretacea. 1899. Hennig: Faunan i Skanes yngre Krita. III. Korallerne. Bihang til K. Sv. Vet. Akad. Handlingar. Bd. 24. Afd. IV. No. 8. Stockholm. 1899. S. 21. PJ. 2. Figs. 27—40. Although this species has not yet been found in the Danien of Denmark, it is included here because it is known from the closely adjacent Swedish Danien locality, Limhamn. I have examined three specimens from the collection in the University of Lund, but as all three showed thick epithecum and a large root extension they belong to the Epitrochus genus and not to the Ceratotrochus. They recall somewhat E. pusillus, for the shell is preserved, (calcareous spar, not arragonite) but are readily distinguished from that form by their greater size and rapid increase in breadth during growth. Locality: Danien. Younger Danien. Coral chalk, Limhamn. 10. Epitrochus pusillus: Forchhammer and Steenstrup. Table III. Figs. 7a—g. Table II. Figs. 5—13. Monomyces pusillus: Steenstrup and Forchammer. 1866. v. Fischer Benzon: Ueber das relative Alter des Faxekalkes. Kiel 1867, S. 17. — M.U.H. 1867. Lundgren: List of fossil Fauna of Sweden. III. Mesozoic, Stockholm, 1888. The species is the most common one found in the chalk pits at Faxe, appearing both among the varieties in the bryozo chalk and in the coral chalk (Ravn’s Naese). It is well- known from earlier times. Numerous specimens are to be seen in the Mineralogical and Zoological Museums at Copenhagen. 220 22 The coral is small. The largest known specimens measure about 15 mm from base to edge of calyx and have a diameter of about 4 mm. It is almost obovate-conical in form, increasing slightly in thickness towards the edge of the calyx. The coral was fastened to the bottom of the sea, or to solid bodies on the bottom by means of a stem and a flat, spreading, basic part, which on its under side shows the im- pression of the object to which it was attached. The upper side of the basic extension is either smooth or has faint longitudinal lines. The usual form, then, is nearly obovate-conical, but seldom entirely regular as the extension often changes resulting in an irregularly bent form. Sometimes growth ceases suddenly and the interseptal compartments are closed with a tabular-like formation. A new individual then grows forth from some portion of the calyx, possibly at right angles to the original extension. The theca is thick and solid, covered with a smooth epithecum which in rare instances permits the costae to shine through. Sometimes horizontal folds are observed. Septa are almost constantly present to the amount of 48, arranged in 5 ordines of 4 cycles. 6 of the first ordo are larger than the others and are easily recognizable in the calyx. Together with the 6 of the second ordo they form, by means of their inner edges, a pseudo-columella which reaches up to the bottom of the calyx and are readily observed here as slightly ruffled chalk leaves. The edge of the septum does not protrude beyond the edge of the calyx; it is not entire but forms a few large patches in the calyx. Further down it billows from one side to another and unites with the neighboring septal edges. In this way the spongy pseudo-columella is formed. The sides of the septa are covered with rows of granulations which are sometimes large and sharp enough to form thorns. These thorns do not unite with similar growths from adjacent septa, but form, especially on the lower sections of the coral, scattered dissepiments which close the interseptal compartments. Locality: Danien: Younger Danien: Coral chalk from Faxe and Aggersborggaard; bryozoa chalk from Faxe. 11. Trochocyathus hemisphaericus, n. sp. Table II. Figs. 14—15. This species is found in cerithium chalk from Stevns Cliff. Both stone kernels and impressions are known. The theca is almost hemispherical, it was once free. Smooth costae have protruded from the outer surface broken by very pronounced rings of growth. The stone kernel is 15 mm in diameter, 10 mm long. It shows traces of 60 septa of which the 6 primary were the largest. The septal edge has not extended beyond the thecal edge, but has penetrated almost immediately into the rather shallow calyx, where it joined the columella. The sides of the septa were covered with rows of prominent granulations. 23 221 There were no dissepiments. No account can be given of the central parts of the coral. Judging by the appearance of the stone kernels there was a true fasciculate Columella at the base of the coral. Whether pali were found and, if so, of what kind, is unknown. In spite of insufficient data about important parts of the coral, I do not hesitate to place it in the Trochocyathus genus on account of its external resemblance to certain French and American corals belonging to that group, — for instance T. apliensis, Fromentel, and T. Hyatti Vaughan. Locality: Upper senone. Cerithium chalk at Stevns Cliff. 12. Brachycyathus parvus, n. sp. Table II. Figs. 16—18. In the coral chalk at Faxe may be found some stone kernels of a single coral which has left no impressions so that it is impossible to describe its outward appearance. The stone kernels, however, are very characteristic and easily recognizable, so I have ventured to determine them. Only small stone kernels are found, the largest specimen known measures 8 mm in height with a diameter of 8 mm at the edge of the calyx. It is regular, obovate-conical in form, somewhat truncated towards the bottom. The stone kernel shows traces of 48 septa, 5 ordines of 4 complete cycles. A true colu- mella has developed from the base causing characteristic grooves in the stone kernel. A cross-cut downward from the edge of the calyx shows that the septa soon reached the center. The calyx was shallow. Septa of the first, second and in part of the third ordo were large, whereas the others only grew out a short distance from the edge of the calyx. Remains of 6 pali which penetrated far into the calyx were found, and a probable fasci- culate columella (table II, fig. 18). Based on these data it seemed right to determine the species as belonging to the Brachy- cyathus genus, in spite of the fact that information.about the epithecum is lacking. Locality: Younger Danien: Coral chalk at Faxe. 13. Caryophyllia danica, n. sp. Table II. Figs. 19—20. In the coral chalk at Faxe single specimens of a Caryophyllia are found. This coral is about 14 mm long; at the edge of the calyx, which is oval, it is 10 mm at the greatest, and 8 mm at the smallest diameter. The form is obovate-conical. The pointed end is broken off so that its mode of adhesion cannot be determined. The theca is smooth with no protuberances and no epithecum. The costae are arranged in longitudinal, slightly granular rows. The calyx is rather shallow. There are about 48 septa, 4 cycles in 5 ordines. The first 2 ordines show equal development and are large, the 222 ; 24 others are less developed. The septal edge is entire. The lateral surfaces of the septa are nearly smooth in so far as they may be seen in the calyx. The columella is covered by a hardened deposit of chalk and cannot be described more closely. Opposite the septa of the third ordo and toward the center are 12 well-developed pali of which the upper edge is entire and smooth and reaches into the calyx to about the same height as the septa. Locality: Danien: Younger Danien: Coral chalk at Faxe. 14. Coelosmilia excavata, v. Hagenow sp. Table III. Figs. 8, 8a, 8b. Table. II. Figs. 21—28. Turbinolia excavata. 1839. v. Hagenow. Monomyces — 1850. Forchhammer and Steenstrup. Turbinolia — 1851. Puggaard: Möens Geologi p. 66. Fig. 9. Coelosmilia excavata 1856. Milne, Ed. et Haime. Histoire naturelle des Coralliaires T. II. 179; — 1858. Fromentel, E. de. Introduction à l'étude des polypes. p. 102. Parasmilia excavata. 1908. Ravn: Faunalisten i Milthers: Kortbladet Stevns og Faxe. D. GA i Række No AK The species is well-known from olden times and belongs to the more common fossils from our white chalk where it is found in its natural condition with the parts of the skeleton preserved and as flint stone-kernels. It is rather common, too, in moraine deposits in gravel beds in secondary layer. It is readily recognized on account of its size, enormous for a single coral. Its determination has been rather difficult. Its inner construction is hard to recognize on account of deposits of solid masses of chalk very difficult to remove without injury to the septa and theca. The original determination of the species as Turbinolia was changed in 1856 to Coelosmilia because dissepiments and the lack of columella were observed. It is plain enough for the same reason that it cannot belong to the Parasmilia genus. The coral is a single coral adhering at the base to other solid bodies at the bottom of the ocean. Its base widens, forming a disc showing on its under surface traces of the object to which the coral was fastened. On the upper surface of this basic disc are fine longitudinal stripes. A thin stem rises from the base, bearing the obovate-conical coral which rapidly increases in breadth; later on the growth in breadth is minimal, and long cylindrical parts are formed in the upper end of which is the calyx (table II, fig. 26). The entire coral may measure about 100 mm from the base to the edge of the calyx, with a diameter of about 40 mm at that point. The theca is rather thin and brittle; it is practically smooth. On the outer surface all the extremities of the septa are visible like slightly protruding costae. When examined through a magnifying glass they are found to be granulated. Countless cross lines, sometimes 25 223 very close together, indicate the pauses in the growth of the coral. There is no true epithecum, but at the point where stem, base and coral merge into each other, a granular epithecum - is found which covers the costae and increases the thickness of the walls and its power of resistance. The number of septa varies from 36—60 in well-developed specimens. These have 4 complete cycles and a fifth incomplete divided into 6 ordines. Of these the 3 first ordines, about 24 in number, penetrate to the center of the coral, while the later septa are weakly and protrude but slightly from the thecal edge. As far into the calyx as one can see after all mud has been removed, no trace of a central formation (Columella) is to be found. A cross-cut of the lowest portions of the coral shows that the septal edges increase somewhat in size towards the center, forming a wavy edge which merges into the neighboring septal edges and forms a fairly regular pseudo-columella, quite independent of the base of the coral, for it is totally lacking at the lowest section of the coral (table II, fig. 27). A few dissepiments are scattered in the interseptal compartments but otherwise there are no endothecal formations. The septa, at all events those of the first ordines, project from the thecal edge and, before the edge vanishes in the bottom of the calyx, form a curve whose edge is entire (table IT, fig. 25). The lateral surfaces of the septa are practically smooth, the small granulated pro- tuberances are, as usual, arranged in curves (table II, fig. 25). Stone kernels of flint are rather common from both Stevns and Möens Cliff and in secondary strata in gravel deposits of the quaternary age. These are, as a rule, spheroidical in shape and consist of a flint kernel with septa preserved as lime septa. No columella is seen in these stone kernels either (table II, fig. 28). Locality: Senone: The zone with belemnitella mucronata: Möen, Stevns. Cerithium chalk. Stevns Cliff. 15. Coelosmilia ponderosa, (Forchhammer and Steenstrup). Table IV. Figs. 1—2. Under this determination we find in the Zoological Museum a few specimens of a Coelos- milia which is closely related to C. excavata. It differs from that form in the great thickness of its theca, its lack of adhesive surfaces and the fineness of the longitudinal stripes on the outer side of the theca. The specimen is incomplete, only the lower part of the coral being present. The entire calyx is wanting. Locality; Senone: The zone with belemnitella mucronata: Aalborg. 16. Coelosmilia brevis, Forchhammer and Steenstrup. Table III. Figs. 2—3. Table IV, Figs. 3—4. In the coral chalk at Faxe are found a few (5) stone kernels of a peculiar shape, like a low cylinder with rounded end surface. The cylindrical portion may be so abbreviated that D. K. D. Vidensk. Selsk. Skr., naturv. og mathem, Afd., 8. Række, V. 3. 29 224 26 the entire stone kernel has about the form of a hemisphere. One shown in table III, fig. 3 is 18 mm in height and 22 mm in diameter at the edge of the calvx. The stone kernel shows traces of about 60—64 septa (4 complete cycles and 1 incom- plete). Only the septa of the 3 first ordines have reached any considerable development, the others were small and very little conspicuous. The lower part of the stone kernels shows indications of dissepiments, for sometimes the single stone kernels in the interseptal com- partments are truncated at different heights so that the ending of the stone kernels is oblique. The ending of the stone kernels indicates lack of Columella but that the septal edges in the lowest parts of the coral were contiguous. A cut through a single stone kernel reveals a deep calyx and contiguous edges of the septa only in the depth of the calyx. The cut which extends to within about 8 mm of the end of the stone kernel shows no traces of Columella. Unfortunately no complete impression of the entire outer side of the coral exists, so it is impossible to state whether it was free or adhered. However, there are impressions of fragments of the coral which show that its costae were well-developed and dentate by means of crosswise furrows. No material was attainable for making a longitudinal cut, so it is impossible to describe the septal edges and the lateral surfaces of the septum more closely. A stone kernel similar in form but flatter was found at Saltholm and should in my opinion be classified as this species. Locality: Danien: Younger Danien: Coral chalk. Faxe, Saltholm. 17. Parasmilia biseriata, M. U. H. Table IV. Figs. 5—9. This species is known from cerithium chalk at Stevns Cliff, but has never been found with its lime skeleton in a state of preservation. Only stone kernels and the impressions of the outer surface of the theca are found. It attains a length of 23 mm and has a diameter at the edge of the calyx of about 8 mm. The impressions show that the form of the coral was obovate-conical with a bent axis. No adhesive extension or stem by which the coral had been fastened was observed. It was free. The stone kernels are short and obovate-conical, often slightly bent; they do not fill out the impression in more than the two upper thirds. The filling of the interseptal com- partments ends at various heights, thus giving the stone-kernel an irregular, obliquely truncate appearance. This is probably due to the fact that dissepiments at different heights have prevented the chalk precipitate from penetrating to the bottom of the interseptal compartments. The theca was thin, either quite smooth on its outer surface or covered with longitudinal stripes of small granulations without any actual costae formation. The septa did not extend beyond the edge of the calyx; its edge was entire and it soon bent down into the rather deep calyx. The sides were granulated with the small protuberances arranged in oblique rows. The medial septal edges were zig-zag, bent, and joined to the pseudo-columella. 27 225 There were about 32 septa in 1 incomplete and 3 complete cycles. Only the septa of the 2 first ordines reached the center, the others protruded but slightly from the edges of the theca. This is known from the stone kernel by the fact that the interseptal fillings lie adjacent, 2 and 2, with deep furrows between the separate groups and slight furrows between the fillings in the same group. This gives the stone kernel a two-serried (bi-seriate) appearance. Locality: Senone. The zone with Scaphites constrictus: Cerithium chalk at Stevns Cliff. 18. Parasmilia cincta, n. sp. Table IV. Figs. 10—12. From the cerithium chalk from Stevns Cliff we have a couple of stone kernels and the impression of a coral, so characteristic that they permit determination. The impressions are very long; the longest is 28 mm with a diameter of about 8 mm. The impressions show that the coral was long, shaped like a worm with many bends. The theca was thin, covered on its outer surface with quite prominent costae, broken by numerous epithecal lines running in a transverse direction, thus dividing the outer surface of the coral into many small quadrilaterals. No actual smooth epithecal covering was seen. The fragments found were not sufficient to determine whether the coral was attached or free. The stone kernels show traces of dissepiments, for they are short, and, measured by the impressions, only a fraction of the entire length of the coral. There were about 30 septa, that is to say, 4 ordines in 1 incomplete and 3 complete cycles. The lateral surfaces of the septum were covered with rows of sharp granulations. Locality: Senone. The zone with Scaphites constrictus. Cerithium chalk at Stevns Cliff. 19, Parasmilia parva, n. sp. Table IV. Figs. 13—16. This little species is easily recognised by its external appearance. It seems to be common to all the Danien strata. Its form is obovate-conical, seen best in young specimens. When older it shows an inclination to irregularly curved elongations. It adheres by means of a small root extension. The largest specimen known measures 15 mm in length and has a diameter at the edge of the calyx of about 5 mm. The majority of the specimens are much smaller. The theca is dotted on its outer surface with thin costae. These are sometimes broken off, forming a row of small ridges. The calyx is deep. The septa protrude but slightly beyond the edge of the calyx and, as a rule, close to the thecal tissue, they sink deep into the bottom of the calyx. Here their inner edges unite, forming a pseudo columella the upper portion of which is just visible. The septal edge is wavy. The lateral surfaces of the septum are covered with rows of quite conspicuous granulations. 29* 226 28 There are 24 septa, that is 3 ordines or 3 cycles. Of these the two first form columella, while the latter are but slightly developed and only protrude a little from the edge of the theca. In the lowest parts of the coral many dissepiments are seen, but no other endothecal for- mations. Locality: Danien. Older Danien: Stevns Cliff; Kagstrup. Younger Danien: Faxe, Rejstrup. 20. Parasmilia cylindrica, n. sp. Table II. Figs. 29—30. I have a great rarity from the white chalk from Stevns Cliff, — the remains of a single specimen of a coral which permits determination in spite of very insufficient material. The fragment of the coral which was obovate-conical in form, with very gradual increase in breadth, so that the fragment is almost cylindrical, is 12 mm in ines with a slightly bent axis. The diameter varies between 3—4 mm. The theca is thin, without epithecum. Costae are present as very low, sharp ridges with protruding granulations here and there. The interstices between the costae are smooth. The material at hand shows no calyx. There are 24 septa; 3 ordines or 3 cycles of which the 12 in the first and second ordines reach the center where they participate in the formation of a spongy pseudo-columella of considerable size. In the interseptal compartments, especially in the lower parts of the coral, scattered dissepiments are found which seal the interseptal compartments. Locality: Senone. The zone with belemnitella mucronata. White chalk, Stevns Cliff. 21. Parasmilia danica, n. sp. Table IV. Figs. 17—18. Only a single defective specimen of this species exists. The coral is 22 mm in length with a diameter of 16 mm at the edge of the calyx. It is obovate-conical with a straight axis. It seems to have adhered directly without any actual basic disc. The theca is thin, without epithecum. Costae are wanting. The calyx is shallow, shows at the edge 48 septa, that is 5 ordines in 4 cycles. Of these the 3 first ordines connect with the large central columella which seems to be fasciculate, while the septa of the 2 last ordines are but slightly developed. The upper edge of the septa protrudes beyond the edge of the calyx. The lateral surfaces of the septa are covered with many rows of well-developed granul- ations. Locality: Danien: Younger Danien. Bryozoa chalk at Faxe. 22. Parasmilia Lindstrômi, Hennig. Parasmilia Lindstrômi. 1899. Hennig. Faunan i Skanes yngre Krita. III. Korallerne. Bihang til K. Sv. Vet. Akad. Handlingar. Band 24. Afd. IV. Nr. 8. S. 15. pl. 2. Figs. 18—33. 29 227 This species is included because HENNIG states that it is also found at Faxe. However, I have not been able to identify it with any form known in the Danish Danien. As I have been unable to obtain specimens from Sweden for comparison I have nothing to add to HENNIG’s description. Locality: Danien. Younger Danien. Limhamn. Faxe. 23. Parasmilia Scanica, Hennig. Parasmilia scanica. 1899. Hennig. Faunan i Skanes yngre Krita. III. Korallerne. Bihang till K. Sv. Vet. Akad. Handl. Bd. 24. Afd. IV. No. 8. S. 20. Pl. 2. Figs. 34—36. I have been unable to find any specimens of this species in the Danish deposits, nor have I been able to see any of the Swedish specimens. Locality: Danien. Younger Danien. Limhamn. 24. Flabellum calcitrapa, v. Koenen. Trochocyathus calcitrapa. 1885. v. Koenen. Ueber eine paleocäne Fauna von Kopenhagen. Göttingen. S. 105. Table V. Figs. 9 a—i. As the now very complete and well-preserved material at hand permits us to see that no pali exist and that the theca is covered by the epithecum which forms slightly curving elongations of the longitudinal axis of the coral, the coral being bent, no doubt exists but that the species in question is a Flabellum and not a Trochocyathus. Nothing further is to be added to v. Koenen’s description. Locality: Danien. Younger Danien. Saltholm. Greensand: Vestre Gasverk and Southern Harbor, Copenhagen. 25. Smilotrochus faxöensis, Forchhammer and Steenstrup. Table III. Fig. 1, 1a. Table IV. Figs. 19—22. The species belongs to those more commonly found in the coral chalk at Faxe. It attains quite a considerable size, up to 30 mm in length, with an average diameter at the edge of the calyx of about 15 mm. It is obovate-conical in form, increasing quickly in breadth, with, as a rule, a strongly bent axis. The theca is quite thin with no epithecum of importance. The costa are but slightly conspicuous, however, opposite the 12 first septa they are plainly visible. The outer surface is covered with numerous fine granulations irregularly scattered over the entire surface. No basic extension was found and no stem. There are 36—48 septa, 3 complete cycles and 1 incomplete one in ordines 1—4. The septal edge protrudes slightly from the edge of the calyx, then, forming a curve, goes back into the calyx. The edge is entire. There is no columella, but deep down in the coral the free septal 228 30 edges grow thicker and unite without, however, forming a central growth deserving of the name columella. As a rule the 6 primary septa are found to be the best developed, but the 6 of the se- condary order are often just as large and can not be distinguished from those of the pri- mary. The remaining septa are small and undeveloped, projecting but slightly from the edge of the calyx. The lateral surfaces of the septa are covered with numerous granulations arranged in the usual curved rows. Toward the inner edge of each septum the granulations are larger and those from adjacent septa may touch each other without merging. No dissepiments or other endothecal formations are seen. The interseptal compartments are always open in their entirety. The stone kernel is very often seen and, on account of the absence of columella, solid and lasting. It is easily recognized by the 48 septal impressions of which every second one penetrates deep into the stone kernel while the alternating one is but slight (table IV, fig. 24). Locality: Danien: Younger Danien: Coral chalk at Faxe. Limhamn and Aggersborg- gaard. 26. Rhizotrochus crassus, Forchhammer and Steenstrup Table III. Fig. 4. Table IV. Fig. 25. In the Zoological Museum is a stone kernel with corresponding impressions from the coral chalk at Faxe, which is a great rarity; nothing similar to it is to be found in the col- lection. As there is only the single specimen, further investigations by means of cutting and polishing were impossible. The stone kernel is 30 mm in length with a diameter at the edge of the calyx of 18 mm. It is obovate-conical in form and regular, though the lower part of the axis is slightly bent. The stone kernel shows traces of about 60 septa, — 4 complete cycles and the fifth incomplete (6 ordines). The impressions left on the stone kernel by the interseptal com- partments show the same grouping in 4’s with deep furrows between as in the species Cera- totrochus ambiguus. There were 12—16 larger septa, the others being small. The lateral surfaces of the septa were strongly granular like the inner side of the theca. The interseptal fillings have therefore a strange prickly appearance characteristic of the species and plainly visible. Table III, fig. 4. The impression of the coral in the chalk shows that it has possessed a thick, smooth epithecum, forming toward the bottom a large basic disc. The upper part of the epithecum was smooth, the lower part and the basic disc, granular. Table IV, fig. 25. There were no dissepiments, and as judging from the lowest section of the stone-kernel there was no columella, it seems justifiable to me to classify the specimen as Rhizotrochus species. Locality: Danien: Younger Danien. Coral chalk, Faxe. 31 229 27. Amfihelia, Becki, n. sp. Table IV. Figs. 26—32. Oculina 1867. Johnstrup: Om Faxekalken ved Annetorp i Skaane. Oversigt o. d. K. D. Vid. Selsk. Forhandl. f. 1866. S. 9. Oculina sp. 1888. Lundgreen: List of the fossil faunas of Sweden. III. Mesozoic, p. 7. In HENNIG’s1) revision of the corals in the Danien deposits in Skaane in which he likewise mentions the forms from Faxe, we read in his introduction: “By Oculina sp. was meant a Lobopsammia or Dendrophyllia species which a secondary deposit of lime has given small calyx-like protuberances scattered over the stem.” In this HENNIG was mistaken. In the material from the Mineralogical Museum in Copenhagen examined by HENNIG there is a very easily recognized and characteristic coral form, which for many years has borne the label, “Oculina.” However, after an examination was made of the recently acquired and well-preserved material (RAvN’s Naese) we find that it is rightly determined only in so far as it belongs to the species, Amfihelia, closely related to Oculina. In honor of the man who more than half a century ago studied these corals, and in many instances in spite of the insufficient material then at hand had the correct conception of them, I am giving this form his name and calling it Amfihelia Becki. The coral forms ramifying colonies of considerable extension. The stems do not increase particularly in size, older and younger stems being of practically the same size. Ramification occurs by gemmation, a new individual growing out from the edge of the calyx of the term- inal individual. This new individual grows a little in length and under the edge of its calyx another new individual in turn grows forth. Irregularities sometimes occur, 2 or 3 individuals growing out of one calyx edge. Where two branches are contiguous, they continue to grow close together, forming a kind of net-work. The calyx is not deep. The theca is solid, thickly coated with a granulated epithecum. The costae protrude but slightly like rows of prickles. The septa hardly extend beyond the edge of the calyx; in the upper parts of the calyx they protrude but slightly beyond the theca; at the bottom of the calyx they touch the rather conspicuous, fasciculate columella. In certain of the calyces there appear to be formations at the inner end of the septa which seem to indicate pali. The septal edge is whole. The septa are almost constantly present to the number of 27 (3 cycles, 3 ordines), of which the 2 first ordines reach the center, while the septa of the third ordo have, as a rule, no connection with the columella. A lengthwise cut through the stems shows that close under the calyx very compact endothecal formations are found which make the stems massive like those in other Oculina forms. Locality: Danien: Younger Danien. Coral chalk, Faxe, Limhamn. 1) A. Hennig: Faunan i Skanes yngre Krita. III. Korallerne. Anhang till K. Sv. Vet. Akad. Hand- lingar. Bd. 24. Afd. IV. No. 8. 1899. S. 5. CONTENTS Flee NONE re TEE a RE een € ~ The species grouped! according ito localities eP-PERPRECR CCR EE ECC ECC CE EE eee . schematic list of the: species and their 2localtiesegr re o ein ta sitet + The separate ‘Species 12 nr. Kelvin Al no RE tot bi: alin PS EON Be rata OR E EEE RE 1. \Haplophyllia faxensis,: Beck PP EEE CEE CEE METEO CEE CEE CT CO CRE ee TRIERER EEE Dendrophyllia ‘candelabrum} Een ee ERP SEC CR CP PCR CR EE Sphenotrochus ‘granulatus; Mo Sp) „ge... NRC TC PE CE SphenO(TOChUS MALUS AV AN CEDENE CEE CE CETTE CIE da eugene hating Ps NE VO ETES Geratotrochus ambiguus, Forchh and Steens Bess SEES REE — Saltholmensis; N.:.SP. HS. ate MANN CR RARE RE io toe oat ER. RBB RUC ÆREN — Milthersii,? niyspis 355515 SC TER BERETTE feed aoe Epitrochus vermiformis,, SP meer eee cdot sped ote ET = supra-eretacea, Hlenmig, 22 2e eee el eee knee = pusillus, Forchh., amd Steenst "enterrer meer Er TO CE Trochocyathus-hHemisphaericus, AL SD EE EE nee ee CT CR CE CCE EEE Brachycyathus parvus, MiiSp. NSG NE MER tien bie MR ee At i] ee tance tal one EEE CERTES Caryophiyllia’ danica;En Sp Pr 2 Slot Be Bee ER ER EE TT CRT EEE Coelosmilia -excavatas sys blag eno Wares tise cer Me Er ST eee PR ET om c — ponderosa; Forehh: and Steenst ete eee TC CETTE EEE — brevis; Forchh. ‚and'Steenst.. "#0 cr. ee Re CR RTE RE RES Parasmiliay biseriata, Borchh. and) Steenst et... RP RE CRE PRET RE — cineta, n. ‘Sp. 2.2 ed ae NP PE RE EE = PATVA, DSPs 2... 008 cielo eckardede ne Robie lke ek ahe CNE ET ee eee — eylindriea; «n::Sp. - .4 or 6408.20 Petra 20 ee ar: EA eet — GAMICA, En: SD: Me mes ce se da eee CE ee bot eae eel ET CU Eee — Lindströmij, Hennig... .-- 2.8 es ee BE ee ae — scanica, Hennig... su... een sectes cities AT wea ee ee Flabellum‘calcitrapas SV Ko0eneD EEE MMM EC POSTS ME SE RE SPORE Smilotrochus'faxoensis, Forchhs/andtSteenst SNS TE SE Rhizotrochus.crassus, Forchh:.and'Sfeenst, PMR = spp ete ee TE 27. Amifihelia: Becki, omy Sp: 202 245 als joys center al SE CN RE TERRIER © œ 1 PURE ww D D ND ND ND Ww ND RRR RR RRR Hi PAS RNR SSHNAAREWN ES EXPLANATION OF TABLE I 1. Calamophyllia faxensis. Pieces of branch !/,. Faxe. 2 — — Calyces */,. Faxe. 3 — — Cross-section of branch ?/,. Faxe. 4. — — Costae ?/,. Faxe. — 5. — — Cross-section of branch ?/,. Faxe. 6 — — Lengthwise section of branch %/,. Faxe. 7 — — Lengthwise section with calyx ?/,. Faxe. 8. Dendrophyllia candelabrum. Lengthwise section with calyx 3/,. Faxe. 9. — — Costae 3/,. Faxe. —= ie — x — Costae °/,. Faxe. — {he — — Tangential cut 3/,. Faxe. — 12. — — Mode of growth !/,. Faxe. — 13. — — Porosity in theca. 3/,. Faxe. — 14. — — Young colony ?/,. Faxe. — 15. _ — Calyx °/,. Faxe. — 16. — — Cross section of branch °/,. Faxe. — 17—20. — — Modes of ramification */,. Faxe. — Pile — — Cross-section of stone kernels ®/,. Faxe. — 22. — — Stone kernels. Surfaces 3/,. Faxé. — 23. Sphenotrochus granulatus. Outer surface of the coral ?/,. Faxe. — 24. — — Calyx "/,. Faxe. — 25. — — Lengthwise section. Columella ?/,. Faxe. — 26. Ceratotrochus saltholmensis. Form of coral 4/,. Saltholm. ile — ; — Form of coral ?/,. Saltholm. — 28. — — Cross-section ?/,. Saltholm. — 29. — — Outer surface °/,;. Saltholm. 530: — — Lengthwise section ?/,. Saltholm. = il — — Septa 4/,. Saltholm. The originals belong to the Mineralogical Museum of the University of Copenhagen. EXPLANATION OF TABLE II Fig. 1—2. Epitrochus vermiformis. ?/,. 1. Outer surface. 2. Cross section. Bulbjerg. pee Oats — — 3/,. 3. Outer surface. 4. Cross section. Stevns Cliff. — 5—7. Epitrochus pusillus. 4/,. Form of the coral. Faxe. == th — — 4/,. Highly developed costae. Faxe. — /9—10. — — 5/,, -2' Calyces. Faxe. silat — — ‘/,. Lengthwise section. Faxe. ur — — 4/,;. Tangential section. Faxe. == 13% — — §/,. Young coral on hydro-coral. Faxe. — KE Trochocyathus hemisphaericus. ?/,. Stone kernels. Cerithium chalk. == 15} — — 2. Outer surface from wax impression. Cerithium chalk. D. K. D, Vidensk. Selsk. Skr., naturv. og mathem. Afd., 8. Række, V. 3. 30 232 34 Fig.16. Brachycyathus parvus. %/,. Stone kernels. Base. Faxe. — 17. — — 3/,. Stone kernels. Outer Surface. Faxe. — 18. — — 3/,. Stone kernels. Cross-section calyx. Faxe. — 19—20. Caryophyllia danica. ®/,. 19 Outer surface. 20. calyx. Faxe. — 21. Coelosmilia excavata. 1/,. Outer surface. Moen. — 22—23. — — ?/,. 22. Outer surface. 23. Cross section. Stevns. — 24. — — 2/,. Lengthwise section. Stevns. — 235. == == 3/,. Upper portion of the septum. Stevns. — 26. — — 1/,. Extended coral. Aalborg. — 27—28. — — ?/,. Stone kernels in flint. Møen. — 29—30. Parasmilia cylindrica. ?/,. 29. Outer surface. 30. Cross-section. Stevns. The originals of Nos. 16—18, 21 and 26 belong to the Zoological Museum of the University of Copen- hagen, the remainder to the Mineralogical Museum. EXPLANATION OF TABLE III Fig. 1. Smilotrochus faxgensis. !/,. 2 stone kernels. Faxe. — lai. = = 1/,. 2 stone kernels. Faxe. — 1b. Ceratotrochus ambiguus. 1/,. Stone kernels. Faxe. — 2—3. Coelosmilia brevis. 1/, Stone kernels. Faxe. — 4. Rhizotrochus crassus. */,. Stone kernels with well preserved impressions showing the large basic disc. Faxe. : — 4a. — — ?/). Impression of the lateral surfaces of the septum. Faxe. — 5—6. Ceratotrochus ambiguus 1/,. Stone kernels. Faxe. — 6a. — — 1/,. Stone kernels cross-section. Faxe. — 7a,b,c,d,e. Epitrochus pusillus. 1/,. Corals of var. forms & sizes. Faxe. — 7f. = — 3), Calyx. Faxe. — 78. — — 3/. Lengthwise section. Faxe. —8, 8 a. Coelosmilia excavata. '/,. 2 corals. Moen. — 8b. — — 1/,. Lengthwise section. Moen. Fig. 8a. has been used by Puggaard as Fig. 9 in his Geology of Moen, p. 66. Under the original tables is written: Monomyces Ehrb. 1. M. faxoensis n. 4. M. crassus n. M. pusillus. ds 2—3. M. brevis. n. 5. M. elongatus Schloth. 8. M. excavata v. Hagen. i 6. M. ambiguus. All the originals belong to the Zoological Museum of the University of Copenhagen. EXPLANATION OF TABLE IV Fig. 1—2. Coelosmilia ponderosa. 1/,. 1. Outer surface. 2. Cross section. Aalborg. — 3 = brevis. 1/,. Section of Stone kernel. Faxe. — 4. — — ?*),. Fractional part of an impression. Faxe. — 5 Parasmilia biseriata. ®/,. Stone kernels and impressions. Cerithium chalk. — 6. — — */,. Impression. Cerithium chalk. — 7. — — 3. Wax model. Cerithium chalk. 8. — — 2/,. Stone kernels. Cerithium chalk. — +95 — — 4/,. Cross-section of stone kernels. Cerithium chalk. 35 Fig.10. ge =, — 13—14. — 15—16. — 17—18. — 19. — 20—21. — 22. — 23—24. — 25. — 26. — 27. — 28. — 29—30. — 31. — 32. 233 Parasmilia cincta. */,. Wax model. Cerithium chalk. — — 3/,. Stone kernels. Cerithium chalk. — — 1/,. Impressions. Cerithium chalk. Parasmilia parva. °/,. 13. Outer surface. 14. Calyx. — — /,. 15. Outer surface. 16. Dissepiments. Parasmilia danica. ?/,. 17. Lengthwise section. 18. Calyx. Faxe. Smilotrochus faxgensis. */,. Outer surface. Faxe. — — 4/,. 20. Outer surface. 21. Calyx. Faxe. — — 3/,. Lengthwise section. Faxe. — 3/,. Stone kernels. 24. Cross-section. Faxe. Rhizotrochus crassus. ?/,. Wax impression of basic disc. Faxe. Amfihelia Becki. °/,. Piece of branch. Faxe. — — 4/,. Piece of branch. Faxe. -. — 3/,. Lengthwise section. Faxe. — — 3/,. Piece of branch. Faxe. — — 4/,. Cross-section just below calyx. Faxe. — — 4/,. Cross-section in calyx. Faxe. The originals of Nos. 1—2, 3, 4, 25 belong to the Zoological Museum of the University of Copenhagen, the others to the Mineralogical Museum. GL 23 | ed ia BE x ki = #3 Stee u | St. Hentze del. Fototypi. Pacht & Crones Eftf. a4 Tab. II EEE Ta ce > Ferner Beer a) | | | St. Hentze del. ‘Fototypi. Pacht & Crones Eftf. Vid Selsk. Skrifter VRække Naturoog math Aa LB. 6 forchhammer oyJ Steenstrup. Gad Danica Tab. UT E fortling Ach: Era.Darertzer-£ C* lath Trst - fo | AMlonomypees Chl? 1 M. faxoensis. n. 4. M. crassus. n. Ÿ M. pusillus. n. 2-3. M. brevis. n. 5. M. dongatus. Sıhloth. 8 M excovatas. vo Hagen. FI 6 M ambiguus. w 3 4 D. K. D. Vidensk. Selsk. Skr., naturv. og math. Afd., 8 R. V. 3. (K. Brünnich Nielsen). Tab. IV St. Hentze del. Fototypi. Pacht & Crones Eftf. on ~ : x cz 3 2 ~ Section des Sciences, 8me série, t. V, n° 1. émoires de l’Académie. Royale des Sciences et des Lettres de Danemark, Copenhague, | DIE RHODANIDE DES GOLDES UND DAS FREIE RHODAN 1 MIT EINEM ANHANG ÜBER DAS GOLDCHLORID VON D. Kei. DANSKE VIDENSK. SELSK. SKRIFTER, NATURVIDENSK. OG MATHEM. AFD., 8 RÆKKE, V. 1 Ss KOBENHAVN HOVEDKOMMISSIONAR: ANDR. FRED. HOST & SON, KGL. HOF-BOGHANDEL BIANCO LUNOS BOGTRYKKERI 1918 Pris: 3 Kr. 50 Ore SRE LT LL QE SL SS Naturvidenskabelig og mathematisk AIRE EN Aimed 42 Tavler, 188085 2. LR Shay, 200 pads MR EN RE ame ee Pen ER Rem os I. Prytz, K.- Undersøgelser over Lysets Brydning i Dampe og {evans Vædsker. SAD ease = 2. Boas, J.E. V. Studier over Decapodernes Slægtskabsforhold: - Med 7 Tavler. Resume en français. 1880 3. Steenstrup, Jap. Sepiadarium og Idiosepius, to nye Slægter af Sepiernes Familie. Med Bemærkninger om to beslægtede Former Sepioloidea D'Orb. og Spirula Lmk. Med 1 Tavle. Résumé en françafs. 1881 4, "Colding, A. Nogle Undersøgelser over Stormen over Nord- og Mellem-Europa af 12te—14de Novb. 1872 og over den derved fremkaldte Vandflod i Østersøen. Med 23 Planer og Kort. Résumé en français. “1881 5. Boas, d.E. V. Om en fossil Zebra-Form fra Brasiliens Campos. Med et Tillæg om to Arter af Slægten Hippidion.""Med?; Taylor: 1881,:.2. sie. ORNE DER N RTE RE PR Rs Sas a 6. Steen, A. Integration af en lineær Differentialligning af anden Orden. 1882 ........ Kr cs ce \ "oe 7. Krabbe, U. Nye Bidrag til Kundskab om Fuglenes Bændelorme. Med 2 Tavler. 1882 -.......... 8. Hannover, A. Den menneskelige Hjerneskals Bygning ved Anencephalia og ‘Misdannelsens' Forhold til Hjerneskallens Primordialbrusk. Med 2 Tavler. Extrait et explication des planches en frahgais, © 1882 ‘9. —— Den menneskelige Hjerneskals Bygning ved Cyclopia og Misdannelsens, Forhold. til Hjerneskallens Primordialbrusk. Med 3 Tavler. Extrait et explic. des planches en français. 1884 ......... 10. —— Den menneskelige Hjerneskals Bygning ved Synotia og Misdannelsens Forhold til Hjerneskallens Pri- mordialbrusk, Med 1 Tavle. Extrait et explic. des planches en francais, (884%; 4 Sree 11. Lehmann, A. Forsog paa en Forklaring af Synsvinklens ee paa Opfattelsen af Lys, ;og parte ved direkte Syn. Med 1 Tavle. Résumé en francais. 1885 . : .. . . . . . . ... . . . . . .. pe gee XX, 'med;20'Tavler, 1881-86-12 . 5 sr int eee SUP ae ots yo, ©, Er EEE 1. Warming, Eug. Familien Podostemaceae. 1ste Afhandling. Med 6 Tavler. Resume et explic. des planches em français. "18812 nu. male Me a en ere te le TM ne ete Solis ele Me segs = Glee OMe ne ae 3 2, Lorenz, L Om Metallernes Ledningsevne for Varme og Elektricitet. 1881 .....:........ 3. Warming, Eug. Familien Podostemaceae. 2den Afhandling. Med 9 Tavler. Résumé et explie. ‘des ev SÅ EAN EAST ISBERG TPE EU SENE SET, ER Sa] NET N pr... ET, 4. Christensen, Odin. Bidrag til Kundskab om Manganets/Ilter. 1883......... EN ee ‘ 5. Lereas, Ie" Farvespredningens Theori. . 4883... focus! nr Sue ie MEN tela Te enue ES St cones ‚6. Gram, J.P. Undersegelser ang. Mængden af Primtal under en given Grænse. Résumé en français. Cissé! 7. Lorenz, L. Bestemmelse af Kvikselvsejlers elektriske Ledningsmodstande i absolut elektromagnetisk Maal a 1885 i EL SN ae es ane mel RO ED sy SBME SE Te EEE 8. Traustedt, M. P. A. Spolia Atlantica. Bidrag til Kundskab om Salperne. Med 2 Tavler. ne et planches en français. 1885" au 2 2 26 mie ra tie tat tlk Unie AT RER he ee NERE RR 9. Bohr, Chr. Om Iltens Afvigelser fra den Boyle-Mariotteske Loy ved lave Tryk. Med 1 Tavle. 1885 . 10. Undersøgelser over den af Blodfarvestoffet optagne Iitmængde udførte ved Hjælp af et nyt Absorpiles meter. Med :2 Tavler 1886 00.02 (omg oS pays by min tal sie Rene OR SEE REINER 11. Thiele, T.N. Om Definitionerne for Tallet, Talarterne og de tallignende Bestemmelser, 1886 ....... XXX, mei ‘6 "Tayler,; 1885—86 +... Sj LIN: Re D NRC ee Ar EAN ER N ae 1. Zeuthen, H. 6. Keglesnitsleren 1: Oldtiden® 2188541252 whee EN ec re palais oer cutee ont ane 2. Levinsen, G.M.R. Spolia Atlantica. Om nogle pelagiske Annulata. Med 1 Tavle. 1885 ......... Woda 3. Rung, 6. Selyregisitärende meteorologiske Instrumenter. Med -it: Tavie) 185: 1%. rss ANNE a 4. Meinert, Fr. De eucephale Myggelarver. Mod 4 dobb. Tavler. Resume et explic. des planches en fran¢ais. LEBER". U 535 ET ne I ES Bee erie) RE RE LH ea Cn RE ve LV, med 25 Tavler. HS8688 su, a it De Adee a ee EUR 1. Boas, J.E.V. Spolia Atlantica. Bidrag til Pteropodernes Morfologi og Systematik samt til Kundskaben om deres geografiske Udbredelse. Med 8 Tavler. Resume en francais. [SEHE ER ae F4 2. ‘Lehmann, A. Om Anvendelsen af Middelgradationernes Metode paa Lyssansen. Med I Tavle. 1886. 3. Hannover, A. Primordialbrusken og dens Forbening i Truncus og Extremiteter hos Mennesket fer Fod- selen; Extrait en francais. : 1887, :/,:. N. Vale ee Re MSN Re RS 4. Lütken, Chr. Tilleg til «Bidrag til’ Kundskab om Ärterne af Slægten Cyamus Latr. eller Hvallusene*. Med’ 1..Tavle. . Résumé en français”. - 1887044 ka = for ud lok oh ee 5. -—— Fortsatte Bidrag til Kundskab om de arktiske Dybhavs-Tudsefiske, særligt Slægten Himantolophus. Med 1 Tavle... Résumé en: francais.’ 1887 Ben ar lee oe FRE Era ANUS 6. —— Kritiske Studier over nogle Tandhvaler af Slægterne. Zursiops, Orca og Lagenorhynchus. Med 2 Tayler.! Resume en francais. -1887/ NS INSEE Oona ate ieee © ee) eee he ee 7,2 Kgefeed, E.. Studier i Platosoforbindelger: 18882454 ua Dre ea 2 0 RON seer Sas 8. Warming, Eug. Familien Podostemaceae. 3die Afhandling. Med 12 Tavler. Résumé et explic. des STAY en Irancals. + 1888:: LM SNe PEL PR NE UN DEN TAC. MUR TERRA TRO pres V, med 11 Tavler og 1 Kort: 1889—91...... eter 16 fein ewe Go 1. Lütken, Chr, Spolia Atlantica. Bidrag til Kundskab om de tre pelagiske Tandhval-Slægter Steno, Del- phinus og Prodelphinus. Med 1 Tavle og 1 Kort. Résumé en français. 1889............ 2, Valentiner, H. De endelige Transformations-Gruppers Theori. Resume en français. 1889 ...:..... 3. Hansen, H,J. Cirolanidæ et familie monnullæ propinquæ Musei Hauniensis. Et Bidrag til cine qu > i -om nogle Familier af isopode Krebsdyr. Med 10 Kobbertavler. Résumé en Santee. 1890. Det Ka Danske Videnskabernes … 6% Rækker : Lorena, L. Analytiske Undersøgelser over a hit 1891 . Selskabs SEERE (Fortsættes pan Daivlagatt SÆDE: ‘ \ a (Fortsat fra Omslagets $, 2.) DER ER PS tee fers tete Gio. tas leren, Pleura og Aortas Væg og Sammensmeltningen deraf » samt de saakaldte Weberske Knoglers Morfolggi. Med mere TAC, Seo LS SAS” She! abel Pen pene de te Ala tu Si ASS Siig Sr ML! re" A lan Odd ta Me es ‘at's fa el oa" SE Me qu hes BT numeriske Funktioner, Resume en frangalas 1890... vor an hehe es, € » Særlig Rotationstiders, Udmaaling. En experimental Undersøgelse. Med 16 Figurer SUS A ANY: USING en a Fm © CL etersen, Emil. Om nogle Grundstollers allotrope Tilstandsformer. 1891 Kone OR a 263 EURO ONCE NE. te alias oh eee “rare ie, Oa! aha, lee leds)" ete ® i 34 Grupper. Résumé et explication des figures en français. 1891 PORES SLL MeL en te hi Mes Sh plea CN SD ots Va) gl aimee es ve eg RE kütken, Chr. Spolia Atlantica. Scopelini Musei Zoologiei Universitatis Hauniensis. Bidrag til Kundskab om det aabne Havs Laxesild eller Seopeliner. Med 3 Tavler. Resume en francais. 1892 Om den elektrolytiske Dissociationsvarme af nogle Syrer. 1892 Bidrag til Seitamineernes Anatomi. Resume en frangais. 1893 F. Sideorganerne hos Scarabæ - Larverne. Les organes latéraux des larves des Searabés. Med Résumé et explication des planches en frangais. 1895 EEF EI EIER ame Bh 6 Ao” mite tal Sy FO ON A Plone Aa eh Aa ene us seat se ‘ En mathematisk Undersøgelse af, hvorvidt Vædsker og deres Dampe kunne have en fælles _ Tilstandsligning, baseret paa en kortfattet Fremstilling af Varmetheoriens Hovedsætninger. Résumé PA RAE STEVE Where le che « God ee Tiere, » daher Sle List TE: up, Japetus, og Lütken, Chr, Spolia Atlantica. Bidrag til Kundskab om Klump- ellér Maanelisken Med 4 Tavler og en Del Xylogralier og Fotogravurer.. 1898 .......... wi 40 8. ale Die edaille belønnet Prisafhandling. Med en Tavle. 1899................... 5 RATS M. Om Zeise’s Platosemiæthylen- og Cossa’s Platosemiamminsalte. Med 1 Tavle. 1900 . tenseu, A. Om Overbromider af Chinaalkaloider. 1900 ............. Se eee FEES IN strup, Japetus. Heteroteuthis Gray, med Bemærkninger om Rossia-Sepiola-Familien i Almindelighed. Re SER TR i ce ee LEDES ne Bille. Om Proteinkornene hos oliegivende Fre. Med 4 Tavler. Resume en français. 1901 . .. "Meinert, Fr. Vandkalvelarverne (Larve Dytiscidarum). Med 6 Tävler. Résumé en français. 1901 Bar Taler, 1899 002,0, ee Din Juel, €. Indledning i Læren om de grafiske” Kurver. Résumé en francais. 809 ENT rt. Imann, Einar, Bidrag til de organiske Kvægsolvforbindelsers Kemi. 1901............... RUE Samsoe Lund og Rostrup, E. Marktidselen (Cirsium arvense). En Monografi. Med 4 Tavler: Résumé en M D A Se En ENE NR DÅ] N Seek eee A ristensen, A. Om Bromderivater af Chinaalkaloiderne og om de gennem disse-dannede brintfattigere For- MEL EE Fa A OBERSTEN SEES CR ae ER foment fais ME pee dene Ae ER AE et Sea ee RS are la ane nat a 0 sur où aire! tase) aie ou La «Mea 6 «a de) ee $ UEIEB en Bing’ 236) N EN En Bhd ona CC SET A . € ristenseh, À. Om Chinaalkaloidernes Dibromadditionsprodukter og om Forbindelser af Alkaloidernes _ Chlorhydrater med højere Metalchlorider. 1904....... OR Ne ee a estes teks 7 Warming, Eug. Familien Podostemaceae. 4de Afhandling. Med c. 185 mest af Forfatteren tegnede Figurer , Kr. 13. udgivne af Det Kgl. Danske Videhskabeines | Seta i (udenfor Skrifternes 6. Rakke, se Omslagets > 29: \ ~ PR 2 Barfoed, €. T. Nogle Undersøgelser over de isomeriske Tinsyrer. 67/........... NE IN; 3 i A Christiansen, (. Magnetiske Undersøgelser. 76.:......-:..-......... Seg ean NW 3 et Ne Colding, A. Undersøgelser om de almindelige Naturkræfter og deres gjensidige Afhængighed, samt: Om Mage | wi netens Indvirkning paa blødt Jern Med 4 Tavler. 50 ........... feo ays ‘4 : RICE —— Undersøgelser over Vanddampene og deres bevægende Kraft i Dampmaskinen. , toa . —— Om Lovene for Vandets Bevægelse i lukkede Ledninger. Med 3 Tavler. 57 ... LS ù —— De frie Vandspejlsformer i Ledninger med konstant Vandfering. Med 3 Tavler. — Om Udstremning af Varme fra Ledninger for varmt Vand. 64 . :. : . .. .. .. tie EK —— Om Stremningsforholdene i almindelige Ledninger ‘og i Havet. Med 3 Tavler. —— Om Lovene for Vandets Bevegelse i Jorden. Med 2 Tavler. Résumé en franc. -—— Fremstilling af Resultaterne af nogle Undersøgelser over de ved Vindens Kraft fremkaldte Stromninger. jHavet.: 276-500 oh ste out delete DO ON SAR SES CR PRE ES este Max Jorgensen, 8. M. Nogle Analogier mellem Platin og Tin. JO HE RL ae PACA NE oid 300 ba ake hy i : —— Om den saakaldte Herapathit og lignende Acidperjodider. 75 ..... rr Si si > Spey z pala Lorenz, L. Experimentale og theoretiske Undersøgelser over Legemernes Brydningstorhold. 12.692 Fae ES ido UNAS LL A oP OR vee RENT À SPAS SI SE KS ahaa). po hee CRUE ARE Norgaard. Bidrag til Oplysning om de kulsure Magnesiaforbindelser. Med 1 Tavle. 50 .. ae bee loa Scharling,’ E: A. © Undersøgelser ‚over Urin.: “43: : We. CRETE ee en EN RUE Aaah SANS ish ed oo —— Undersegelser over den Qvantitet Kulstof, som i Form af Kulsyre forlader det menneskelige Legeme bac de ca. Dagnets Leb A302. ii CN EL Be Mey ae mye eet NE ES SR NE Fi Ape FA ee 27 j å —— Fortsatte Forsøg for at bestemme Kulsyren i Menneskets SAD aan dj: Med 1 Tavle. Al eset TES pk —— Tredie "Række af samme. 49 ............... ee Ret Re eet gt RE om ofS ARRET —— Bidrag til Oplysning om de i Handelen forekommende Balsamers kemiske Forhold. GER ER 5 EN —— Om Dbglal og -Athal. 5522, 022. 205 EN) St ee RP ke en REE ER SEERE SAR Thomsen, Jul. Bidrag til et thermochemisk Systems 52 Nec oe apelin at Cenk gh MCE es os ace oe —— Den elektromotoriske Kraft udtrykt i Varmeeenheder. 58. :..... RE er A DE EE RATS u —— Thermochemiske Undersøgelser over _Affinifetsfor rholdene mellem hope, og Baker i ‘vandig ‘Oplasiting 2 APN se ® Med 1 Tavle. Résumé en frang. 69.....::........ FT Ne DE VR OE 9 DANNE er ere FED — 006, V-VUE We SEG JU Da oy IS pk Smee 5 2 DE AS BO. AX, GIO. He kh. ET As Oo ee enema We RIDE AIO TAN AL © MN alle ME À ii VAS ENT 2E $s aus Se AR BES SES REDE gh eee HR Se Ves aie ENS +00. XI "Med ‘en Tävle: 73 Fe halted aes a ae hor DR NE AUS hotte rag RR, | Re CS — 0: AKT 78 ne abate: AR wile NN RS PESTE DIN UC, CURE RÉUNIE EL. LEP Topsøe, H., & Christiansen, C. Krystallografisk-optiske Undersøgelser, med særligt Hensyii til isomorfe Stoffer. 73. Zeise, W. C.° Où Acechlorplatih: ‘4te. .p. VOGT SE BER pou Lita tase Tak ee SPAT Pea ky Ti -— Om et Product af Ammonium-Sulphocyan- Hydrat ved Ghlor?2-43 2,7. 2%... 8 M US « PS SAC JOUE -— Undersøgelser over Producterne ved Tobakkens terre Destillation og over Tobaksragens chemiske Be ut skallenhed; ‘43.47.5808 STONE SALEN Sie dire 10 eut De. a oa SE tee PA ee | RU ss Om Virkningen mellem xanthogensyret Kali og Jode. 45...... DE Na coy inte RU LE ; | — Om Virkningen mellem Kali-Methyloxyd-Sulphocarbonat ag Jode. 47... ........ LA SA * så + Mémoires de PAcadémie Royale des Sciences et des Lettres de Danemark, Copenhague, Section des Sciences, 8me série, t.V , no 2. THE LACTIC ACID BACTERIA | | BY SF ORLA-JENSEN TAR EVT SR ee ERAT | : WITH 51 PLATES | | + D. KGL. Danske VIDENSK. SELSK. SKRIFTER, NATURV, OG MATHEMATISK, AFD., 8. RÆKKE, V. 2 ey) © MES CNP MS ARCO AEG | | | | | (| x £ 3 KOBENHAVN + HOVEDKOMMISSIONÆR: ANDR. FRED. HOST & SON, KGL. HOF-BOGHANDEL : E BIANCO LUNOS BOGTRYKKERI | i | Fy | SRE OE ET Pris (Text & Plates): 46 Kr. Pre on ex 6. . Jorgensen, S.M.: Det kemiske Syrebegrebs Udviklingshistorie indtil 1830. Efterladt ee Det Kgl. Danske Videnskabernes Selskabs Skrifter. Naturvidenskabelig og mathematisk Afdeling, öde Række. É, 1915-1917. eet ene ee eee RER ee 200 VE ee A ee SON AR papers tore Prytz, K. og J. N. Nielsen: Undersøgelser til Fremstilling af Normaler i Metersystemet, grundet paa Sammenligning med de danske Rigsprototyper for Kilogrammet og Meteren. 1915.......... Rasmussen, Hans Baggesgaard: Om Bestemmelse af Nikotin i Tobak og Tobaksextrakter. En kritisk »Undersagelse. “19167... RER eh tees lee eas oe oes et eee SEE WE Christiansen, M.: Bakterier af Tyfus-Coligruppen, forekommende i Tarmen hos sunde Spæd- kalve og ved disses Tarminfektioner. Sammenlignende Undersøgelser. 1916 ..............-..+. Juel, C.: Die elementare Ringfläche vierter Ordnung. 1916 ...... LEE RS SE Zeuthen, H. G.: Hvorledes Mathematiken i Tiden fra Platon til Euklid blev én rationel Viden- skab. Avec un résumé en frangais. 1917 Se RS aa en ae ea GS N hae = IK, med:4 ‘Tavier, 1916-1918, 7.2 me er PL SR coh os Rebeca ane EEE udgivet af Ove. Jorgensen og "SPS, Sorensen. TOU alts ER RC ee ER 2 Hansen-Ostenfeld, Carl: De danske Farvandes Plankton »i Aarene 1898—1901. Phytoplankton og Protozoer.. 2. Protozoer; Organismer med usikker Stilling; Parasiter i Phytoplanktonter. Med 4 Figurgrupper og 7 Tabeller i Teksten. Avec un résumé en français. 1916 .................:. Jensen, J. L. W. V.: Undersøgelser over en Klasse fundamentale Uligheder i de analytiske. Funk- tioners. Theori. .E- 1916... ua an an tore ences a PRE oe eal a cece RS CIDRE EE PAS ars cee Pedersen, P. O.: Om Poulsen-Buen og dens Teori. En Experimentalundersogelse. Med 4 Tav- ler: 1917.52, nn M atic ey eA tN, hor Pei oe ES Ci Re Le te TD ET i Juel, C.: Die gewundenen Kurven vom Maximalindex auf einer Regelfläche zweiter Ordnung. 1917 Warming, Eug.: Om Jordudlobere. With a Résumé in English 1918 ...................... III, med. 14 Kort’og 12° Tavler, 19172 191977 ee = ce ARR Eee Wesenberg-Lund, C.: Furesostudier. En batlıymetrisk Undersogelse af Molleaaens Seer. Under Medvirkning af Oberst M. J. Sand, Mag. J. Boye Petersen, Fru A. Seidelin Raunkiær og Mag. sc. C. M. Sleenberg. Med 7 bathymetriske Kort, 7 Vegetationskort, 8 Tavler og ca. 50 i Texten trykte Figurer. Avec un résumé en | francais. 1907. CAEN TE PE nee ER Se Sl N Lehmann, Alfr.: Stofskifte ved sjælelig Virksomhed. With a Résumé in English. 1918 ...... Kramers, H. A.: Intensities of Spectral Lines. On the application of the Quantum Theory to the problem of the relative intensities of the components of the fine structure and of the stark effect of the lines of the hydrogen spectrum. With 4 plates. 1919.................... ses. IV (under Pressen). Bohr, N.: On the Quantum Theory of Line-Spectra (under Pressen). V (under Pressen). Bjerrum, Niels und Kirschner, Aage: Die Rhodanide des Goldes und das freie Rhodan. Mit einsm Anhang über das!.Goldchlorid 71918 LR nice. sree mere sts ol ik RE EE ES Pr TR CSP Orla-Jensen, S.: The lactic acid Bacteria. With 51 Plates. 19197... 2... 2282 Stat eg OE VI (under Pressen). Christensen, Carl: A Monograph of the genus Dryopteris. Part II (under Pressen). cadémie es des Sciences ei des Lattre de Danemark, ‘Copenhague, — Section des Sciences, 8me série, t.V , no 2. BY \ WITH 51 PLATES MED Kor. DANSKE VIDENSK. SELSK. SKRIFTER, NATURV. OG MATHEMATISK AFD., 8. RÆKKE, V. 2 PLATES ae’ KOBENHAVN | | HOVEDKOMMISSIONÆR: ANDR. FRED. HØST & SON, KGL. HOF-BOGHANDEL Ae - BIANCO LUNOS BOGTRYKKERI 1919 7 i ; x ; ers Pris (Text & Plates): 46 Kr. i Valles {oi > L RC. > wei. te 14 1. © won . Jørgensen, S.M.: Det kemiske Syrebegrebs Udviklingshistorie indtil 1830. Efterladt Manuskript, . udgivet ‘af Ove Jorgensen og SEPT Sarensen. 1916 or. CR ee eee NS _the problem of the relative intensities of the components of the fine structure and of the stark 2 7 einem Anhang über, das" Goldehloridk 19132 ig wc. 0. clare sires SS EUS SK SEER RESEN “See AOS u + , e, ae £ SENT ES ae : 44", Dee (mae ee > ser OK ae Det Kgl. Danske Videnskabernes Selskabs Skrifter Naturvidenskabelig og mathematisk Afdeling, - 8de Række. RE, 1915-1977 ee Me ene foes fag Bis ee. ons AR een > Prytz, K. og J. N. Nielsen: Undersogelser til Fremstilling af Normaler i Metersystemet, grundet paa Sammenligning med de danske Rigsprototyper for Kilogrammet og Meteren. 1915:....... 0 Rasmussen, Hans Baggesgaard: Om Bestemmelse af Nikotin i Tobak og Tobaksextrakter. En kritisk Undersogelse. 1916 ........ RE RE MA MRC ARE A ET Re: ene Pee ES Christiansen, M.: Bakterier af Tyfus-Coligruppen, forekommende i Tarmen hos sunde Spæd- kalve og ved disses Tarminfektioner. Sammenlignende Undersøgelser. 1916 ............1....., Juel, C.: Die elementare Ringfläche vierter Ordnung. 1916 ........... oP RP Te er ee Zeuthen, H. G.: Hvorledes Mathematiken i Tiden fra Platon til Euklid blev en rationel Viden- skab.: Avec.un_résumé en -frangais, Or Re pew ci cee pests a ee le FORK II, med 4 Tavler, 1916—19385 I RER EE Hansen-Ostenfeld, Carl: De danske Farvandes Plankton i Aarené 1898—1901. Phytoplankton og Protozoer. 2. Protozoer; Organismer med usikker Stilling; Parasiter i Phytoplanktonter. ae 4 Figurgrupper og 7 Tabeller i Teksten. Avec un résumé en frangais. 1916 ................ are Jensen, J. L. W. V.: Undersogelser over en Klasse fundamentale Uligheder i de analytiske Funk- tioners" Theori: I? 1936)... Soe. sae eee RP ies tae TOT aR eee Oe ee : Pedersen, P. O.: Om Poulsen-Buen og dens Teori. En Experimentalundersogelse. Med 4 Tay- ler. .1917 Fe. sl ir eee de seat PR Maca ace eeu EE by g pea EE BR este Juel, C.: Die gewundenen Kurven vom Maximalindex auf einer Regelfläche zweiter Ordnung. 1917 Warming, Eug.: Om Jordudlobere. With a Résumé in English LIST TPE EME TRE III, med 14 Kort og 12 Tavler, 1917—1919 ......... OMS aot Ve ae ee Wesenberg-Lund, C.: Furesostudier. En bathymetrisk Undersogelse af Molleaaens Soer. Under pr af Oberst M. J. Sand, Mag. J. Boye Petersen, Fru A. Seidelin Raunkier og Mag. sc. ; M. Steenberg. Med 7 bathymetriske Kort, 7 Vegetationskort, 8 Tavler og ca. 50 i Texten trykte Figurer! Avec un résameé en franeaiss 197% rar, ane eos wget RE hate) rele CR TR SRE CR Lehmann, Alfr.: Stofskifte ved sjelelig Virksomhed. With a Résumé in English. 1918 ee Kramers, H. A.: Intensities of Spectral Lines. On the application of the Quantum Theory to. effect of the lines of the hydrogen spectrum. With 4 plates. 1919................... ete eee . . 11 J IV (under Pressen). Bohr, N.: On the Quantum Theory of Line-Spectra (under Pressen). . x V (under Pressen). Bjerrum, Niels und Kirschner, Aage: Die Rhodanide des Goldes und das freie Rhodan’ Mit Orla-Jensen, S.: The lactic acid Bacteria. With 51 Plates. 1919 .........,....... rears: NE VI (under Pressen). Christensen, Carl: A Monograph of the genus Dryopteris. Part II (under Pressen). À Section des Sciences, 8™* série t.V, n° 3. E- 4 Mémoires de l'Académie Royale des Sciences et des Lettres de Danemark, Copenhague, ZOANTHARIA FROM SENONE AND PALEOCENE DEPOSITS IN DENMARK AND SKAANE | BY K. BRUNNICH NIELSEN WITH 4 PLATES D. Ker. DANSKE VIDENSK. SELSK. SKRIFTER, NATURV. OG MATHEMATISK AFD., 8. RÆKKE, V. 3. +--+ KOBENHAVN HOVEDKOMMISSIONÆR: ANDR. FRED. HOST & SON, KGL. HOF-BOGHANDEL BIANCO LUNOS BOGTRYKKERI 1922 Pris: Kr. 5,25. 1. jørgensen, S.M.: Det kemiske Syrebegrebs Udviklingshistorie indtil 1830. Efterladt Manuskript,- TEE ae RER Det Kgl. Danske Videviakahernes Selskabs Skrifter. Naturvidenskabelig og mathematisk Afdeling, öde Reekke. I, L915 1917. ae er. ee ee ee le ee Ne ES CEE Prytz, K. og J. N. Nielsen: Undersogelser til Fremstilling af Normaler i Metersystemet, grundet paa Sammenligning med de danske Rigsprototyper for Kilogrammet og Meteren. 1915.......... Rasmussen, Hans Baggesgaard: Om Bestemmelse af Nikotin i Tobak og Tobaksextrakter. En kritisk Undersogelse.. 1916 u... 0:8... 0.5 „een ae > el er Re RE ES EC EE Christiansen, M.: Bakterier af Tyfus-Coligruppen, forekommende i Tarmen hos sunde Spæd- kalve og ved disses Tarminfektioner. Sammenlignende Undersøgelser. 1916 ................... . Juel, C.: Die elementare Ringflache vierter Ordnung. 1916 7 2 RENE ME a ere eee Zeuthen, H. G.: Hvorledes Mathematiken i Tiden fra Platon til Euklid blev en rationel Viden- skab. Avec un résumé en francais: -1917....272 ws EE A EC IX, med 4 Tavler, 1916—1918 3.2.5. Ha... ee ders mae eee EE udgivet af Ove Jorgensen og’ S. P. L.. Sørensens 191622... 2 ea ela ee EEE Hansen-Ostenfeld, Carl: De danske Farvandes Plankton i Aarene 1898—1901. Phytoplankton og Protozoer. 2 Protozoer; Organismer med usikker Stilling; Parasiter i Phytoplanktonter. Med 4 Figurgrupper og 7 Tabeller i Teksten. Avec un résumé en français. 1916 .........:........ . Jensen, J. L. W. V.: Undersogelser over en Klasse fundamentale Uligheder i de analytiske Funk- tioners ‘Theori, I." 1916. ::....:,04 0000 CR a ape EC CR EE . Pedersen, P. O.: Om Poulsen-Buen og dens Teori. En Experimentalundersogelse. Med 4 Tav- Fer: SUD pace CRE MN RE RE PE PCR ED RS Be rn REN . Juel, C.: Die gewundenen Kurven vom Maximalindex auf einer Regelfläche zweiter Ordnung. 1917 Warming, Eug.: Om Jordudlobere. With a Résumé in English. 1918 . ar ae re ne egy ee ee III, med 14 Kort 08:12 Tavler; 19171913. EN SEER Wesenberg-Lund, C.: Furesostudier. En batlıymetrisk Undersogelse af Molleaaens Soer. Under Medvirkning af Oberst M. J. Sand, Mag. J. Boye Petersen, Fru A. Seidelin Raunkier og Mag. sc. C. M. Steenberg. Med 7 bathymetriske Kort, 7 Vegetationskort, 8 Tavler og ca. 50 i Texten trykte Figurer: Avec un résumé en francais: 1917 oc 2. wat EE ete 21s hes bol ae CC RE TE Lehmann, Alfr.: Stofskifte ved sjælelig Virksomhed. With a Résumé in English. 1918 ...... Kramers, H. A.: Intensities of Spectral Lines. On the application of the Quantum Theory to the problem of the relative intensities of the components of the fine structure and of the stark effect of the lines of the hydrogen spectrum: "With 4 plates» 1913772 Ze IV (under Pressen). Bohr, N.: On the Quantum Theory of Line-Speetra. Part I. 19182 ee — Samme, Part. II. 19185... 20. ee SS ee ee ee er Se re V (under Pressen). Bjerrum, Niels und Kirschner, Aage: Die Rhodanide des Goldes und das freie Rhodan. Mit einem Anhang über das Goldchlorid: . 1918.% SSG SAS ee ern SEES SES as Re - Orla-Jensen, S.: The lactic acid Bacteria. With 51 Plates; 1919 2 0. een K. Brünnich Nielsen: Zoantharia from Senone and Paleocene Deposits in Denmark and Skaane. With 4 Plates. 1922 Ce ee ee wr ewe er wm CC re er ar ar er ac BRS KER Er Sr Zr u Er Er Er Zr er WI (under Pressen). Christensen, Carl: A Monograph of the genus Dryopteris. Part II. 1920.................... Lundblad, O.: Süsswasseracarinen aus Dänemark. Mit 12 Tafeln und 34 Figuren im Text. 1920. VII (under Pressen). Wesenberg-Lund, C.: Contributions to the Biology of the Danish Culicidæ. With 21 Plates and 19 Figures in the text. 1920—21 NN 100205714