short chains of a few links, very often in pairs; usually actively motile because of flagella located at the ends. Endospores known in only two varieties.

1. Cells short, slightly bent, rigid, comma-like, sometimes attached in a screw-like manner, always one (exceptionally two) flagellum at the end. According to Hüppe, they possess arthrospores. Vibrio1 O. F. Müller, emend.

Löffler.

2. Cells long, spirally bent, like a corkscrew, rigid, usually with a polar bunch of flagella formed of many long principal and several short accessory ones. In the Spir. sputigenum Miller the bunch of flagella is not at the end, but on the side. Spirillum Ehrenb., emend. Löffler.

3. Cells consist of flexible, long, spiral, coiling threads. Flagella unknown. Locomotion by means of an undulating membrane is suspected. Spirochete Ehrenb.

In a strict sense the causes of glanders, diphtheria, leprosy, and actinomycosis do not belong among the fission-fungi. It is generally acknowledged to-day that they must either be designated as fission-fungi, which form a connecting-link to the higher fungi (hyphomycetes), or

1 Migula, with Schröter, called the group which is now almost universally designated as vibrio, microspira—a designation that is unnecessary if we accept the definition of vibrio suggested by Löffler. Moreover, Schröter's definition of spirillum and microspira does not suit the known properties of the varieties therein included. For the few non-motile (without flagella) rigid vibriones Migula has introduced the name Spirosoma Migula.

2 Here belong such forms as the Spirillum endoparagogicum Sor., described by Sorokin and which he once found in a hollow tree in Kasan. This remarkable typically spiral-shaped organism formed typical endospores, which germinate while still within the spirillum, and so offer characteristic pictures (C. B. 1, 466). The organism appears to connect the spirillacea and the bacilli. According to Prazmowski, the Vibrio rugula possesses a spore causing swelling of the end where it is located. Spore-formation has not been described in other vibriones. We know nothing regarding the flagella of this vibrio rugula, which resembles the Bac. oedematis maligni. Moreover, Zettnow expressly contradicts the idea that the vibrio rugula forms

spores.

1

candidly as low hyphomycetes, as was done for the first time in the first edition of this book, in 1896. Kruse has placed the actinomyces, together with its nearest relatives, in a family of hyphomycetes, "streptotricheæ," while he still speaks of a Bacillus tuberculosis, etc. Recently, Lachner-Sandoval has introduced the name actinomycetes to designate the group of "fission-fungi closely related to the hyphomycetes" (as we had designated them in the first edition), and until we have something better it answers for practical purposes.

SUPPLEMENT I.

Actinomycetes (Lachner-Sandoval).

Delicately threaded organisms, free of chlorophyll, with true branching, in part very abundantly ramifying mycelium, partly with the formation of conidia. Young cultures often present only unbranched rods resembling bacteria, which can in no way be differentiated from ordinary fission-fungi. According to many authors there is a tendency to the formation of clubs or knobs at the ends of the threads.

1. Microscopic: Slender often somewhat bent rods, often with a tendency to a clubbed swelling of the ends, branches rarely observed in young cultures, easily broken off, and often difficult to find also in old cultures. Always nonmotile; never conidia.

a. Rods stain interruptedly (striped) with weak staining-solutions, since the organism is composed of parts with different staining properties. Not stained by the method

1As hyphomycetes there have been designated for a long time in botany a large number of threaded fungi, of which nothing is known except threads and non-sexual spores that are upon threads or special carriers. The group is constantly growing smaller, as many earlier "hyphomycetes" have become known as members of the sharply characterized groups of fungi (ascomycetes, zygomycetes, basidiomycetes). The actinomycetes appear to form an entire natural group of the "hyphomycetes."

for the tubercle bacillus. Clubbed, wedge-shaped, and pointed rods frequent. Corynebacterium L. and N.

3. Rods stain with usual staining-solutions with difficulty or generally not at all. Stain by the tubercle bacillus method, i. e., it is acid resisting. Clubbed swelling of the ends in cultures rare, in tissues somewhat more often. Mycobacterium L. and N. 1

2. Mycelial threads, long, thin, extended, or winding, without dividing partitions, with delicate sheaths and true branches. Many species separate from the air-hyphæ rows of short spores (conidia), which, whitish and moldlike, project upward above the solid nutrient substratum; in connection with other species, conidia-formation is unknown. Not stained by tubercle bacillus method. Motility sometimes absent, sometimes present. Almost all varieties emit a musty odor. Actinomyces Harz.

over

We have determined to follow the example of Gasparini and designate this genus as actinomyces. Streptothrix, as these varieties, together with others, are called by Kruse, is a name given by Corda in 1839 to a certain mold-like organism which Cohn, because of an sight, in 1875 introduced a second time into the literature. Cladothrix, which many authors to-day call these varieties (compare Günther), is the designation for an entirely dif ferent pseudodichotomous plant (see Supplement II). the first edition we accepted, with Sauvageau and Radais, the old designation of Wallroth, oospora, but since Lach

In

1 Since we proposed this name in the first edition, we have seen that Metschnikoff (Virchow's Archiv, 113, p. 70, 1888), who first recognized the peculiar position of the tubercle bacillus as opposed to the other then known bacteria, in a work "Regarding the Phagocytic Rôle of the Tubercular Giant Cell," has said: "If one considers that in the perfected stage the tubercle bacteria have grown into (although short) threads, and also differ from other analogous forms (except the lepra bacteria) in the possession of a very dense envelope, then perhaps it will be easier to accept the designation Sclerothrix for the genus, and Sclerothrix Kochii for the species of the tubercle bacterium." We should have immediately accepted these names if we had known of them, but believe that according to the rules of botanical nomenclature our names should now stand, since Metschnikoff only made a conditional proposal, did not accurately define his new genus, and never made any use of the new name himself, while we have ourselves already established a name.

ner-Sandoval (Dissert. Strassburg, 1898) has convinced us that the true oospora varieties are much larger although similarly constructed organisms, we also, with this author, consider the name actinomyces (Harz) at present the most correct.

Some varieties of wide practical importance, closely related to bacteria, but reminding one very strongly of true algi (oscillaria), have been included under Supplement II.

If we cast a glance over this system, we can not deny that the families and genera are often connected by transition varieties; we recall only the following: The border between the coccaceae and bacteriacea is obliterated by oval and lance-formed (!) cocci and certain extremely short bacilli (compare, in the special part, Micr. melitensis, Bacterium Fraenkelii); between streptococcus and micrococcus, micrococcus and sarcina, it is often unsafe to distinguish. In the cycle of growth of many bacilli twisted forms occur; flagella and endospores occur in such various forms that it would lead to an entirely unnatural grouping if the attempt were made to found a classification that depended in part upon the flagella or endospores.

The Bacterium Fraenkelii Hashimoto, for which we are indebted to the kindness of the authors, unfortunately died before we could study it.

Upon solid nutrient media the organism forms short rods with polar flagella; upon fluid media, on the contrary, it forms quite long chains of cocci and occasionally sarcina forms. Thus it connects the coccaceae with the bacteriacea, as does the Micr. melitensis, and shows, as we have indicated above in other examples, that sarcina forms occur as growth forms in cocci and that the presence of flagella is also variable. (See Hashimoto, Z. H. XXXI, 85.)

B. Systematic Description of the Most Important Varieties of Fission-fungi. INTRODUCTORY REMARKS TO THE SYSTEMATIC PART, ABBREVIATIONS, ETC.

1. We have described about eighty species as completely and exhaustively as possible, several hundred are briefly described, and many

varieties which we are not acquainted with in detail are briefly referred to in the connection where they belong.

2. The colonies, slightly magnified, are described and drawn with closed diaphragm, and so placed that the peripheral portions are sharply visible.

3. For the drawings and descriptions plates with a medium number of colonies, 60 to 100, were always employed. Usually the smaller colonies were selected.

4. All statements, unless otherwise qualified, regarding the growth upon gelatin apply at a temperature of 22°, upon agar at 37°.

5. When nothing particular is said regarding the color and consistency in the description of the agar streak culture, and of the surface growth in the agar stab culture, they are the same as upon the agar plate.

6. Regarding the formation of pigment, odoriferous, gustative, and other metabolic products nothing has been said unless special investigations have been made upon the same.

7. Our original purpose of treating exhaustively the resistance of all important varieties to injurious influences has been abandoned as being too far-reaching. This decision was also partially dependent upon the fact that the statements of authors often deviate so widely. Therefore we have restricted ourselves to making complete statements regarding some varieties (Micr. pyogenes, Strept. pyogenes, Strept. lanceolatus, Bac. anthracis, Bact. typhi, Corynebact. diphtheria, Mycobact. tuberculosis, Vibrio cholera).

8. References to the illustrations in the atlas are always given thus: Plates with Arabic numerals, figures with Roman. Thus, 5, VIII, signifies figure VIII in Plate 5.

The introductory remarks of the separate sections, coccaceæ, bacteriaceæ, spirillaceæ, are also to be heeded.

Statement of the Terms Employed by Us in the Description of Cultures of Bacteria.

I. STAB CULTURES.

A.

Not liquefying.

1. Stab canal:

(a) Thread-like: Uniform growth without anything especially

characteristic.

(a) Smooth.

(B) Rough.

(b) Nodular: The stab canal is beset with larger or smaller tubercles, points or teeth.

(c) Hairy: The stab canal is beset with delicate longer or shorter undivided spurs, which are (a) parallel, (3) curled, (y) matted.

(d) Branched: The stab canal is beset with dividing outgrowths. (e) Beaded: The stab canal consists of small roundish or round connected colonies.

(f) Band-like: Growth as a small band, produced by making the stab canal with a loop.