Pigment production appears to be lessened by merely infrequent transfers of the agar culture.

The Bact. prodigiosum at 37° forms no pigment; if grown at this temperature for a long time with constant transfers, the power of pigment formation is, even under favorable circumstances, lost for many generations (Schottelius).

Very interesting experiences with pigment-forming cultures of varieties which usually produce colorless growths are scattered through the literature; for example, Fawitzky regarding yellow to rusty-red colonies of Streptococcus lanceolatus; Kruse and Pasquale on colored varieties of the Streptococcus pyogenes (Ziegler's Beiträge, XII); also the experience recently published by Kutscher, according to which a pseudoglanders bacillus, when obtained from the animal, in the primary culture upon serum develops a bright orange-red growth, but after a few transfers the red color is completely exchanged for a white one (Z. H. XXI, 156). Perhaps still more important is the observation, which is easily made, that, from some internal condition, at times colored and colorless colonies of the same variety grow side by side upon a plate; for example, in Bact. kiliense. R. O. Neumann has, by selection, grown from the Micr. pyogenes a aureus white, yellow, and red varieties (A. H. xxx, 1).

An analogue to this variation from internal causes is related by F. Hildebrand, who observed in a stock of Iris florentina, which always bore pale-blue flowers each year, the sudden appearance of two flowers presenting dark violet portions arranged in sectors (Ber. d. deutsch. bot. Gesell., 1873, 476).

2. The Formation of Ammonia and Urea-fermentation.

According to Sommaruga (Z. H. XII, 273), aerobic bacteria growing in non-saccharine nutrient media always form an alkali from albuminous bodies.

When sugar is present, most varieties, besides producing alkalis, form acids from the sugar. In this way is explained the fact that many young cultures of bacteria at the beginning are neutral or faintly acid in reaction because

of the small amount of sugar in the bouillon (originating in the meat). When the sugar is exhausted, then alkali production advances more strongly (Th. Smith).

The bodies causing the alkaline reaction, so far as known, are ammonia (at times it may be smelt), amine, and ammonium bases. To determine the amount of alkali formed, one titrates single tubes which contain 10 c.c. of peptone bouillon, both uninoculated and one to fourteen days after inoculation, with decinormal acid, using phenolphthalein as indicator. The difference upon titration gives the increase of alkali.

The following may be useful as an example of the production of alkali by bacteria which, in the presence of sugar, form acids energetically (for 100 c. c. amounting to 5 c.c. to 7 c. c. normal acid). One hundred c.c. of a nutrient medium was employed, which contained a trace of meatsugar, and was originally exactly neutral with phenolphthalein.

A special instance of the production of an alkali by bacteria occurs in the transformation of urea into ammonium carbonate, CO(NH2)2+2HO, CO, (NH4)2.

In 1896 we stated that of sixty varieties tested, only the Bact. vulgare, Bact. prodigiosum, and Bact. kiliense were found able to decompose urea. Brodmeier (C. B. xvш, p. 380) has investigated the urea-splitting property quantitatively with the Bact. vulgare, and my pupil, Dr. Mann, has recently done the same with the Micrococcus pyogenes a aureus and albus, with two forms of the colon bacillus, and several sarcinæ. One hundred c. c. of filtered urine, sterilized at 85°, after being ten days in the incubator contained abundant NH,. Mann could not demonstrate any action by the same culture of the Bact. prodigiosum, which we had previously found to cause energetic fermentation of urea. Thus, this property also is variable, and

the contradictory results of authors with the Bact. coli (see special part) and the Micr. pyogenes are thus explained.

What are described in the literature as Micrococcus ureæ Leube, Bacillus ureæ Leube, Bacillus ureæ liquefaciens Flügge, can be partially identified as the Micr. pyogenes y albus and Bacterium coli, but the descriptions of these varieties allow of no accurate identification. The urea-splitting function appears to occur occasionally in very many varieties. Warington (C. B. vi, 498), Burri, Herfeldt and Stutzer (C. B. L. 1, 284) have described ureasplitting varieties. Compare also the investigations of Miquel (Ann. d. Micrographie, Bd. I u. f.), which are very interesting biologically, but which lose much in value because Miquel has elaborated a very singular nomenclature which does not take into consideration the usual varieties. Miquel has observed varieties which are able to decompose as much as 60 gm. of urea to a liter. He claims to have isolated a special ferment, urase, which decomposes

urea.

The older literature can be found as given by Leube (Virchow's Archiv, Bd. c, p. 540); the newer, with the method of determining ammonia (according to Schlösing), by Mann.

3. Formation of Complicated Basic Metabolic

Products.

Especially through the investigations of Brieger (Ueber Ptomaïne, Heft 1-III, Berlin, Hirschwald), besides ammonia, a large number of basic, crystalline, nitrogenous bodies are known as products of bacterial metabolism. These bodies are usually called ptomains (@μa, putrefaction) or putrefaction alkaloids. They occur, so far as closely studied, mostly in the following groups:

1 For a long time the poisonous ptomains were called toxins, yet now most authors call all bacterial poisons toxins, without reference to their chemical constitution, and usually one understands the term to include the "albuminous-like" bacterial poisons more especially.

NH2

amin-putrescin, isomer of sepsin; pentamethylendiamin, called cadaverin, etc. Of these the most poisonous is ethylendiamin.

2. Ammonium bases. The best known is

15

Nearly related are muscarin (C,H15 NO ̧), vinylcholin (CH1 ̧NO), neuridin (C,H,,N2), etc.

3. Pyridin derivatives. Derived from pyridin (CHN), the following are especially found : Collidin (CH11N), parvolin (C,H13N).

4. Indol (C,H,N) and skatol (C,H,N). Compare page 79.

In addition the following are known: Amido-acids (leucin, tyrosin, etc.) related to guanidin, C(NH)(NH2) 2, and also numerous, insufficiently or feebly characteristic bodies, whose enumeration here would be useless, since the poisons among them are not now recognized as the essential disease poisons, as was the case in former years.

The isolation of these bodies can here be only hinted at. The method of Brieger, which is most employed, is as follows: Brief boiling of the culture or "decomposed material " rendered weakly acid with hydrochloric acid; reduction of the filtrate to a syrup; dissolve in 96% alcohol and free from impurities (especially traces of albumin) with lead acetate; removal of the lead; concentration of the filtrate and precipitation from this of the double salt of mercury of the ptomain by means of an alcoholic sublimate solution. After removal of the alcohol by heat and the mercury by hydrogen sulphid, there is produced the characteristic double gold and platinum compound, whose crystalline quality is an index of its purity. One may try to obtain directly the crystalline hydrochlorate, and by the aid of caustic soda the free bases, which are often fluid.

Some ptomains, like very many plant alkaloids, as soon as set free by caustic soda, are easily obtained in aqueous solution with ether. Still, Brieger's procedure is widely useful, as it takes into consideration many bodies

which do not dissolve in ether. For extensive review of literature upon ptomains, see Jacquemart (C. B. ix, 107).

4. Production of Complex “Albumin-like" Poisonous Metabolic Products ("Toxalbumins," Toxins).

In addition to the discussion of the relatively simply formed, basic, less poisonous metabolic products of bacteria, we may speak, as briefly as possible, of the other bacterial poisons. From the standpoint of our present knowledge, they may be divided into three classes:

1. Bacterioplasmin (Buchner). Compare Hahn, Münch. med. Wochenschr., 1897, No. 48, 1344. The expressed juice (compare pp. 29 and 65) of bacteria contains, in ordinary unchanged form, poisonous substances; present in the bacteria of cholera, typhoid, and tuberculosis; absent in Micr. pyogenes and Bact. anthrax. Koch's new tuberculin, "TR," is essentially also a plasmin.

2. Bacterioprotein (Buchner). Under this head are placed albuminous bodies, unaltered by heat, which produce fever (pyrogenic), inflammation, and suppuration1 (phlogogenic). They are obtained by boiling for several hours scraped potato-cultures with 0.5% caustic potash (about 50 volumes KHO to 1 volume of bacterial substance). The protein may be precipitated from the clear fluid obtained by filtering through paper, by carefully rendering it feebly acid. The protein is filtered out, washed and dried, and before being employed is dissolved in a small quantity of weak soda solution.

The best-known protein is the "old" tuberculin of Koch; also mallein belongs here. According to Buchner and Römer, all bacterial proteins act alike and not specifically.

3. "Toxalbumins," now usually called toxins. The isolated statements of earlier investigators (Christmas, Roux and Yersin, Hankin) were confirmed to a great extent by the investigations of C. Fränkel and Brieger (Berl. klin. Wochenschr., 1890, 241 and 268), who found that

1 Suppuration is best produced by bacterial and non-bacterial products if they slowly diffuse from a gelatin capsule into the subcutaneous tissues (Poliakoff, C. B. XVIII, p. 33).