I. Introduction

A. Introductory remark  When I was invited to write this chapter I felt flattered. I have always been interested in names, in etymology and semantics. The invitation was probably due to more than 25 years of active membership in the International Committee for Systematic Bacteriology (ICSB) and in the Editorial Board of the International Journal of Systematic Bacteriology, and there especially my self-adopted task of watching the correctness of new Latin names by offering advice in etymology and questions of procaryote nomenclature. What I write hereafter is an outflow of the experiences I have gathered in these tasks including correspondence in etymological (often intertwined with nomenclatural) matters with hundreds of colleagues. Therefore, I shall try to write this chapter from the viewpoint of the microbiologist—as a user; for the user—rather than writing it ex cathedra as a classicist might want to do. Further, what I write here are my own opinions on these matters and they are not meant to offend anyone who has other or better insights.

B. The Latin/Greek thesaurus of words and word elements  Scientific terminology, both in technical terminology and in nomenclature, has to fulfill requirements other than those of everyday language. These requirements have been excellently described by the late Fritz C. Werner 1972, a German zoologist.

The first requirement is that every term must unambiguously circumscribe a clearly conceivable idea and that every name stands for a special object or a special group of objects characterized by determined features.

The second requirement is that the total number of different words and word combinations must exceed the large number of discernible objects and abstract concepts, thereby ensuring that names are unambiguous. This is a real challenge as the number of objects, processes and concepts is continuously growing both in depth and breadth because of new scientific and social developments, and changes in nature due to human activities.

As more scientists from a wider range of nationalities participate in these developments, it is important that scientific terms and names fulfill a third requirement, namely universal comprehensibility.

These three requirements—unambiguousness, a large number of possible combinations, and universal usage—are met, to a high degree, by the fact that the terminology of natural sciences and medicine is largely based on the lexicon of classical Greek and Latin. The fact that these so called “dead” languages no longer undergo natural and living changes makes their word material a thesaurus that has been used and may be used further for contemporary needs. Consequently one has more or less arbitrarily given these classical words and word elements certain new meanings. Using a living and constantly changing language in this way would promptly lead to problems and misunderstandings.

Firstly, the use of ancient word material allows the naming of the many new and—in their numbers—permanently increasing objects and concepts for which there are no respective words in contemporary spoken languages; even circumscriptions and combinations of words would hardly suffice. Latin and Greek offer a wealth of word elements and ways to form words that remain inexhausted thus far and are likely to serve our needs for a long time in the future, although scientists have not always been careful or reasonable in their “creations”. By mixing Greek and Latin elements, by dropping syllables, repositioning letters, contracting words and creating arbitrary formations, the antique wealth of words has been changed, at times rather significantly. Furthermore, many other languages have contributed, and the names of scientists and other persons have been latinized.

What Werner 1972 did not emphasize was the fact that Latin remained the international language (lingua franca) of philosophy, religion, law, sciences, and politics throughout the European Middle Ages and the Renaissance and for philosophical and scientific publications up into the nineteenth century. Its usage, although limited to these circles, led to an enormous increase in vocabulary, usually adopted from other European or oriental languages (e.g., Arabic). It also needs to be mentioned here that Latin has remained the spoken language in the center of the Catholic Church, the Vatican, and is likely to be so into the future. This is particularly well documented by the fact that the Libraria Editoria Vaticana takes all efforts to integrate new Latin words coined for modern objects and concepts into the written and spoken Latin of the Vatican. The Lexicon Recentis Latinitatis, that appeared 1992 in Italian and 1998 in German, contains about 15,000 new Latin words, “from astronaut to zabaione”, word combinations and circumscriptions of the fields of sciences, technology, religion, medicine, politics, sports, and even common idiomatic terms.

The thesaurus of words, enlarged this way, is thus no longer identical with that of either classical language but represents “something new” that has developed along historical lines and follows special contemporary laws of language.

All of the statements made by Werner 1972 apply to general scientific and medical terminology as well as to biological nomenclature. And they apply especially to the scientific nomenclature of procaryotes (eubacteria and archaebacteria) and viruses because these—in contrast to most animals, plants, and larger fungi—do not have popular or vernacular names in any living language because of their usual invisibility.

Nomenclature (“the system of names used in a branch of learning or activity”) is an indispensable tool for correct information in our fast growing scientific world with its rapidly developing information networks. The binomial nomenclature used in biosystematics goes back to 1735 when the Swedish botanist Carolus Linnaeus (Karl von Linné, 1707–1778, ennobled 1757) published his famous “Systema Naturae” in Latin according to the scholarly habits of his times.

By introducing the species concept and the use of Latin and Greek for the names of living beings, Linnaeus laid down the principles of modern biological systematics as well as nomenclature. In our “age of informatics” one could certainly think of other ways to name the vast number of plants, fungi, animals, protists and procaryotes, perhaps by a number and/or letter code. For the human brain, however, names are still easier to memorize and work with as part of a system, as long as they are readable and pronounceable.

For the scientific names of procaryotes the International Code of Nomenclature of Bacteria (ICNB, Bacteriological Code), issue of 1992, is the compulsory compendium of governing Rules. It is the task of the accompanying chapter on nomenclature by P.H. A. Sneath to explain the Bacteriological Code (ICNB), whereas this chapter is intended to deal with etymology. Etymology means “origin and historical development of a word, as evidenced by study of its basic elements, earliest known use, and changes in form and meaning” or “the semantic derivation and evolution of a word”. “Etymology” is derived from Greek etymon, “the truth” and thus aims at the true, the literal sense of a word.

Etymology is a necessary element in biological nomenclature as it explains the existing (i.e., so far given) names and helps to form new names. For the average microbiologist, “etymology” is that part of a species or genus description that stands first, describes the accentuation, origin and meaning of the name, contains a lot of strange abbreviations and is often considered as superfluous or nasty. I shall come to appropriate examples at the end of the chapter.

In 1993, the late professor of classical languages, Thomas Ozro MacAdoo of Blacksburg, VA, U.S.A., wrote a marvelous chapter on “Nomenclatural literacy” (MacAdoo, 1993) with the intention of helping bacteriologists form correct names. MacAdoo carefully described and examplified the five Latin declensions, the Greek alphabet and its Latin equivalents, the Greek declensions and their Latin equivalents, adjectives and participles, compounding in Latin and Greek, and the latinization of modern proper names. It cannot and will not be my task to equal this excellent and scholarly piece of work, as it contains an introduction to the two classical languages and requires a basic knowledge of, at least, Latin grammar. I highly recommend reading, or better studying, MacAdoo's paper. But I am afraid that I cannot agree with him on the way personal names should be latinized nowadays. (Additional literature recommended as etymological help for the formation of new bacterial names is marked by an asterisk in the further reading list.)

C. Pronunciation and accentuation  For many bacterial names the current common pronunciation differs from the pronunciation that is correct according to Latin rules (cf. common text books for Latin). It is unfortunately strongly influenced by the speaker's mother tongue, a clear indication that Latin is no longer the lingua franca of the scientific world. Whereas native speakers of languages that are written close to phonetics, such as Italian, Spanish, Portuguese, Dutch, or German, usually pronounce Latin close to its spelling, native speakers of French and especially of English (languages pronounced rather differently from their spelling) often pronounce Latin according to the pronunciation rules of their languages, i.e., further away from the written form. These differences in pronunciation are not generally that important as differences in spelling, because the name in question is often understood despite differences in pronunciation. Substantially helpful here, however, could be to pronounce at least the vowels as they are pronounced in Spanish and Italian, languages whose pronunciations stayed close to their Latin origin. International science will have to live with this problem until the day when all languages are written according to phonetic rules.

In many Central European high schools Latin pronunciation has gone back to the times of Caesar and Augustus when the Romans always pronounced the letter c as the sound k. As a consequence students pronounce, e.g., Caesar “Kaesar” (origin of the German word Kaiser which means emperor) or Cicero “Kikero”. In bacteriology this leads to alternate pronunciations of Acinetobacter, Acetobacter, etc. (as akinetobakter, aketobakter, etc.) by some younger European microbiologists.

I consider it a pity that, for scientific terms used mainly in chemistry and physics, the writing of Greek k remained (keratin, kinetics) whereas in biological nomenclature it has usually, but not always, been latinized to c (Triceratops, Acinetobacter). Fortunately, classical Latin already introduced the Greek z for transliterated Greek words, and Medieval Latin introduced the letter j for the consonantic i. Meanwhile several names of bacteria starting with J have been proposed (e.g., Janthinobacter and the specific epithet jejuni). It makes sense to use the j in Latin names as the first letter of a word or word element when it is followed by a vowel.

One significant problem with pronunciation is that of some personal or geographical names used in generic names or specific epithets, e.g., the bacterial generic name Buttiauxella, named after the French microbiologist Buttiaux (pronounced: “buttio”). This generic name and specific epithets like “bordeauxensis”, “leicesterensis”, or “worcesterensis” may be pronounced fully (as Latin would require) or pronounced as though they were spelled “buttioella”, “bordoensis”, “lesterensis”, “woosterensis”. I am afraid that we will have to leave the decision of pronunciation in such cases to the single scientist, as a rule for such “problems” seems rather difficult to conceive.

Frequently accentuation of Latin names appears to pose problems, especially when Greek word elements are involved. Here, the correct classical accentuation is often not used in bacterial names, e.g., the accepted accentuation of the name Pseudomonas is pseu-do-mo'-nas, whereas the classical Greeks would have accentuated the word pseu-do'-mo-nas. An almost universal guideline for accentuation of generic names is, that the syllable next to the last bears the accent. Although this holds for most specific epithets as well, we do tend to encounter other accentuations more often. The practical sense of natural scientists should prevail and the present common usage of accentuation in bacterial names should be the guideline.

II. Formation of generic names and specific epithets

Since Linnaeus, biological species bear binomial names, consisting of a genus (kind) and a species (appearance) name. The latter, if taken by itself, is called “specific epithet”. A complete species name thus consists of the genus name and the specific epithet. In principle the language of biological nomenclatural names is Latin. In nomenclature, words of Greek origin as well as those of any other origin are handled as Latin, i.e., they have to be “latinized”.

Only those bacterial names contained in the Approved Lists of Names (Skerman et al., 1980) and the Validation Lists that regularly appear in the International Journal of Systematic Bacteriology have standing in nomenclature. Regularly updated non-official lists of legitimate bacterial names (except for cyanobacteria described under the Botanical Code) are published by the German Culture Collection DSMZ, Braunschweig, Germany, twice a year. Dr.J.P. Euzéby, Toulouse, France, provides an even more detailed non-official list electronically on the Web site www-sv.cict.fr/bacterio/.

A. Compound names  Compound names are formed by combining two or more words or word elements of Latin and/or Greek origin into one generic name or specific epithet. In most cases two word elements are used (e.g., Thio/bacillus, thio/parus), but up to four elements may be found (e.g., Ecto/thio/rhodo/spira).

In order to avoid later changes, these recommendations (cf. Bacteriological Code, Appendix 9 [Lapage et al., 1992]; Trüper, 1996) should be strictly followed, i.e., they should be considered as rules without exceptions.

The reader may protest here and mention, e.g., Lactobacillus as being against this ruling. Lactibacillus would indeed be the correct name, however, the name Lactobacillus is much older than the Bacteriological Code and has become a well established name. The ending -phile (or -philic) in English is often added to words of Latin origin connected by -o- (e.g., acidophile, francophile, anglophile, nucleophile, lactophile etc.). This is due to the meaning of -phile, “friendly to”, which commands the dative case. In the most common Latin declension, the second, the dative is formed by adding an -o to the stem (acidophile, friendly to whom/what? friendly to acid). Therefore in bacteriology we have a number of older compound names of Latin origin with the connecting vowel -o-. By unknowingly taking over such originally dative-derived word elements ending on -o, names like Lactobacillus came into existence. Such cases prove that Appendix 9 of the Bacteriological Code (Lapage et al., 1992) does not have the power of a Rule yet. In the future new name formations of that kind should be avoided.

There are numerous mistakes with respect to compound names. Sometimes authors want to express that their new organism was isolated from a certain part of an animal's body, e.g., from the throat of a lion; throat is pharynx (Greek word stem: pharyng-), lion is leo (Latin word stem leon-). These stems may be correctly combined in two ways: “pharyngoleonis” or “leonipharyngis”. Unfortunately the authors chose leopharyngis, which may be corrected to the latter. This example demonstrates the different connecting vowels as well. Two more examples may emphasize the importance of word stems: so Obesumbacterium should be corrected to Obesibacterium, as the Latin stem of the first component is obes-, and the connecting vowel must be -i-. The generic name Carbophilus was formed the wrong way, because the stem of the first component is carbon-; the correct name would be Carboniphilus. For those scientists without training in Latin, a good Latin dictionary indicates the genitive of a noun thereby allowing them to identify the stem of a Latin noun. Typically, the genitive usually shows the stem (e.g., carbo, carbonis, the coal) well. MacAdoo 1993 gives a very useful overview on word stems and declensions for non-classicists. An excellent pocket book on word elements (stems) of Latin and Greek origin for usage in scientific terms and names was published by Werner 1972. However, it has only appeared in German to date. An English translation would be of great value for biologists world wide.

Other typical, yet well established misnomers whose connecting vowels were not dropped include Acetoanaerobium, Cupriavidus, Haloanaerobacter, Haloanaerobium, Haloarcula, Pseudoalteromonas, Streptoalloteichus, Thermoactinomyces, Thermoanaerobacter, Thermoanaerobacterium, not to speak of numerous equally malformed specific epithets.

B. Generic names  The name of a genus (or subgenus) is a Latin noun (substantive) in the nominative case. If adjectives or participles are chosen to form generic names they have to be transformed into substantives (nouns) and handled as such.

Both Latin and Greek recognize three genders of nouns: masculine, feminine, and neuter. Adjectives associated with nouns follow these in gender. For the correct formation of specific epithets (as adjectives) it is therefore necessary to know the gender of the genus name or of its last component, respectively.

The more frequent last components in compound generic names of masculine gender are: -arcus, -bacillus, -bacter, -coccus, -ferax, -fex, -ger, -globus, -myces, -oides, -philus, -planes, -sinus, -sipho, -vibrio, and -vorax; of feminine gender: -arcula, -bacca, -cystis, -ella, -ia, -illa, -ina, -musa, -monas, -opsis, -phaga, -pila, rhabdus (sic), -sarcina, -sphaera, -spira, -spina, -spora, -thrix, and -toga; of neuter gender: -bacterium, -bactrum, -baculum, -bium, -filamentum, -filum, -genium, -microbium, -nema, -plasma, -spirillum, -sporangium, and -tomaculum.

Correct examples of these three ways are Staphylococcus aureus (adjective: “golden”), Desulfovibrio gigas (nominative noun: “the giant”), and Escherichia coli (genitive noun: “of the colum/colon”), respectively.

1. Adjectives and participles as specific epithets Latin adjectives belong to the first, second, and third declension. Those of the first and second declension have different endings in the three genders, whereas in the third declension the situation is much more complicated, as there are adjectives that don't change with gender, others that do, and those that are identical in the masculine and feminine gender and different in the neuter. Table 1 gives some representative examples. Note also that comparative adjectives are listed. I recommend always checking an adjective in the dictionary before using it in the formation of a name.

Participles are treated as if they were adjectives, i.e., they fall under Rule 12c, (2), of the Bacteriological Code. Infinitive (also named “present”) participles in the singular do not change with gender. According to the four conjugations of Latin they end on -ans (e.g., vorans devouring, from vorare to devour), -ens (e.g., delens destroying, from delere to destroy, deleo I destroy), -ens (e.g., legens reading, from legere to read, lego I read), -iens (e.g., capiens, from capere to seize, capio I seize), -iens (e.g., audiens, from audire to listen, audio I listen). Note that the knowledge of the ending of the first person singular in the present is decisive!

Perfect participles change their endings with gender and are handled like adjectives of the first and second declension, e.g., voratus, vorata, voratum devoured, deletus, deleta, deletum destroyed, lectus, lecta, lectum (irregular) read, captus, capta, captum (irregular) seized, auditus, audita, auditum, listened/heard.

2. Nominative nouns in apposition as specific epithets While the above mentioned first and third ways to form specific epithets are generally well understood and usually do not pose problems, the formation of epithets as substantives in apposition has obviously been misunderstood in several cases. So, for instance, when the name Mycoplasma leocaptivus was proposed for an isolate from a lion held in captivity, the authors, probably unintentionally, called their bacterium “the captive lion”, whereas they wanted rather to explain the origin of their isolate “from a captive lion”. Thus “captivileonis” would have been the correct epithet.

A nominative noun in apposition does not just mean that any nominative noun may be added to the generic name to automatically become its acceptable epithet. In grammar, apposition means “the placing of a word or expression beside another so that the second explains and has the same grammatical construction as the first”; i.e., the added nominative noun has an explanatory or specifying function for the generic name, like in general English usage “the Conqueror” has for “William” in “William, (called) the Conqueror”. Thus Desulfovibrio gigas may be understood as Desulfovibrio dictus gigas and translated as “Desulfovibrio, called the giant”, which, with reference to the unusual cell size of this species, makes sense.

Because all specific epithets ending with the Latin suffixes -cola (derived from incola, “the inhabitant, dweller”) and -cida (“the killer”) fulfill the above-mentioned requirement, they are to be considered correct.

Most legitimate specific epithets formed in bacteriology as nominative nouns in apposition so far have been mentioned and, where necessary, corrected recently (Trüper and De' Clari, 1997, 1998).

Although they are not explicitly ruled out by the Bacteriological Code, I have not yet encountered tautonyms, i.e., specific epithets identical with and repeating the genus name, in bacterial nomenclature (such as in zoology Canis canis, the dog). In order to avoid confusion, it would be wise to abstain from proposing such names.

3. Genitive nouns as specific epithets The formation of specific epithets as genitive nouns rarely poses problems, as the singular genitive of substantives (nouns) is usually given in the dictionaries.

If the plural genitive is preferred, as, e.g., in Rhizobium leguminosarum (“of legumes”), one has to find out the declension of the noun, as plural genitives are different in different declensions. This question will be addressed below.

D. Formation of bacterial names from personal names  Persons may be honored by using their name in forming a generic name or a specific epithet. This is an old custom in the whole area of biology. The Bacteriological Code, however, strongly recommends to refrain from naming genera (including subgenera) after persons quite unconnected with bacteriology or at least with natural science (Recommendation 10a) and in the case of specific epithets to ensure that, if taken from the name of a person, it recalls the name of one who discovered or described it, or was in some way connected with it (Recommendation 12c).

It is good style to ask the person to be honored by a scientific name for permission (as long as she/he is alive). Authors should refrain from naming bacteria after themselves or co-authors after each other in the same publication, as this is considered immodest by the majority of the scientific community.

The Bacteriological Code provides only two ways to form a generic name from a personal name, either directly or as a diminutive: Both are always in feminine gender.

Appendix 9 of the ICBN recommends how such names should be formed. Appendix 9 has, however, not the power of the Rules.

The application of the classical Roman rules for name-giving, as was done by MacAdoo 1993, does not make sense as modern names worldwide follow different and various rules and regulations. A differentiation in prenomina, nomina, and cognomina is therefore no longer applicable and should not be used as a basis for latinization of names nowadays. Principally, modern family names are either nomina or cognomina in the classical sense. Continuing latinization of names as practiced in ancient Rome would have the advantage that the practice would not change over time. Rather, it would remain fixed. Therefore MacAdoo 1993 would have preferred to establish a uniform rule for latinization of names. But attention must be paid to the fact that since classical times throughout the Middle Ages up into the nineteenth century, (usually learned) people of others than the Roman nation have latinized their names, and thus several varieties of latinization have developed and must be considered as historically evolved. Thus, if such names are not incorrect, they cannot be denied or refused under the Bacteriological Code (Appendix 9). I have therefore tried to give the recommended rulings of Appendix 9 (adopted as editorial policy by the Bergey's Manual Trust) a simpler and clearer wording and have given examples according to those latinizations that have historically precedence (Trüper, 1996). The results were revised and are compiled in Table 2.

Some personal names in Europe were already latinized before 1800 and kept since then. If they end on -us, replace the ending by -a or -ella (diminutive) respectively (e.g., the name Bucerius would result in “Buceria” or “Buceriella”). Beware, however, of Lithuanian names like Didlaukus, Zeikus etc.! These are not latinized but genuine forms and would receive the ending -ia according to Table 2.

No more than one person can be honored in a given generic name or epithet. In the case of the Brazilian microbiologist Henrique da Rocha Lima, the generic name Rochalimaea was formed by dropping the particle da and combining his two family names. Combinations of the names of two or more persons cannot be constructed under this aspect. Here the only possibility would be the provision of the Bacteriological Code for forming “arbitrary names”. These are treated below.

If an organism is named after a person, the name cannot be shortened, e.g., “Wigglesia” after Wigglesworth, “Stackia” after Stackebrandt or “Goodfellia” after Goodfellow etc., but must fully appear. Certainly titles (Sir, Lord, Duke, Baron, Graf, Conte, etc.) and particles (de, da, af, van, von, etc.) indicating nobility or local origin of the family should not be included in bacterial names, although they may belong to the name according to the laws of the respective country.

Rarely, generic names or specific epithets have been formed from forenames (first names, given names, Christian names), i.e., not from the family name, so the genus Erwinia was named after the American microbiologist Erwin F. Smith. The first name Elizabeth appears in Bartonella (formerly Rochalimaea) elizabethae. One could imagine that, in avoiding the usually long Thai family names first names should be chosen in respective cases. Also unusually long double (hyphenated) names like the (hypothetical) Basingstoke-Thistlethwaite or Saporoshnikov-Shindlefrink hopefully do not occur so often among microbiologists as to be honored by a bacterial name (hyphens are not allowed, anyhow!).

This simplified scheme should perhaps be recommended by the Bacteriological Code as an optional alternative to Appendix 9. Such a ruling should, however, not be introduced with retroactive power as Principle 1 of the Bacteriological Code aims at constancy of names.

To form specific epithets from personal names there are, in principle, two possibilities: the adjective form and the genitive noun form. The adjective form has no means of recognizing the sex of the honored person, which, in principle is not necessary for nomenclatural purposes. The personal names receive appropriate endings according to the gender of the generic name as indicated in Table 3. Thus an adjective epithet is formed that has the meaning of “pertaining/belonging to … (the person)”.

When the genitive of a latinized personal name is formed for a specific epithet, the sex of the person to be honored may be taken into consideration as indicated in Table 4.

On the basis of classical, medieval, and modern usage any of the forms of latinization listed in Table 4 may be chosen. As evident from Table 4 the formation of specific epithets from personal names as genitive nouns poses certain problems only with names ending on -a and -o.

Classical Roman names of male persons like Agrippa, Caligula, Caracalla, Galba, Seneca, etc. (predominantly cognomina) were used in the first declension like the masculine nouns poeta (the poet), nauta (the sailor), or agricola (the land dweller, farmer), regardless of the fact that most of the nouns in this declension are of feminine gender. If bacteria would have been named after these gentlemen, their specific epithets were agrippae, caligulae, caracallae, galbae, and senecae, respectively. I think that Volta, Migula, and Komagata are dignified successors in this row.

By now the reader will understand that possibilities 2–5, although permissible or even recommended by MacAdoo 1993, look and sound rather awkward and are likely to produce numerous misspellings. Therefore I strongly suggest to use the classical version and version 1 only.

Roman names ending on -o usually followed the third declension, i.e., the genitive is formed by adding the ending -nis, which also reveals that such words have stems ending on n, e.g., Nero/Neronis, Cicero/Ciceronis, or the noun leo/leonis (the lion). Medieval Latin followed this custom. So, for the medieval German emperors named Otto the genitive Ottonis was used in writing, which was all in Latin at that time. Therefore it makes sense to treat Spanish, Italian, Portuguese, Japanese, Chinese, Ukrainian, Indonesian, as well as all other names that end on -o the same way.

Several European names are derived from classical Greek and end on -as, such as Thomas, Andreas, Aeneas, Cosmas, etc. In their genitive form, they receive the ending -ae: Thomae, Andreae, Aeneae, Cosmae, etc. Although one could argue for a Latinization to Thomasius, Andreasius, etc., to form the specific epithets thomasii, andreasii, etc., I would recommend the use of the classical ending -ae.

E. Formation of bacterial names from geographical names  Authors often consider it necessary to indicate the geographical origin, provenance, or occurrence of their isolates in the respective specific epithets.

Such epithets are simply constructed by adding the ending -ensis (masculine or feminine gender) or -ense (neuter gender) to the geographical name in agreement with the latter's gender. If the name of the locality ends on -a or -e or -en these letters are dropped before adding -ensis/-ense (e.g., jenensis from Jena, hallensis from Halle, bremensis from Bremen). Sometimes authors make the mistake of adding iensis/-iense. This is only correct if the locality's name ends on -ia (e.g., California leads to californiensis). The advice given above guarantees that such mistakes will not happen.

Specific local landscape names such as tundra, taiga, puszta, prairie, jungle (from Sanskrit jangala), steppe and savanna may be dealt with in the same way (tundrensis, taigensis, pusztensis, prairiensis, jangalensis, steppensis and savannensis, respectively).

Epithets on the basis of geographical names may not be formed as substantives in the genitive case, as if they were derived from personal names (e.g., the city of Austin, Texas, cannot lead to “austinii” but must lead to “austinensis”).

Quite a number of localities in the Old World (Europe, Asia, Africa) have classical Greek, Latin, and medieval Latin names and adjectives derived from these: europaeus, aegyplius, africanus, asiaticus, ibericus, italicus, romanus (Rome), germanicus, britannicus, hibernicus (Ireland), indicus (India), arabicus (Arabia), gallicus (France), polonicus, hungaricus, graecus (Greece), hellenicus (Hellas, classical Greece), hispanicus (Spain), rhenanus (Rhineland), frisius (Friesland), saxonicus (Saxony), bavaricus (Bavaria), bretonicus (Brittany), balticus (Baltic Sea), mediterraneus (Mediterranean Sea), etc.

Since the discovery of the other parts of the world by European sailors and travelers, European geographers have continued to give Latin names to “new” continents and countries, so adjectives like americanus, cubanus, mexicanus, etc. were introduced. Wherever older adjectives exist they may be used as specific epithets to indicate geographical origins.

European and Mediterranean cities and places of classical times may have had very different names than those in current useage: e.g., Lucentum (Alicante, Spain), Argentoratum (Strasbourg, France), Lutetium (Paris, France), Traiectum (Utrecht, Netherlands), Ratisbona (Regensburg, Germany), Eboracum (York, U.K.), Londinium (London, U.K.), Hafnia (Copenhagen, Denmark). Microbiologists are free to demonstrate their knowledge of these ancient names but may use epithets derived from the present names of such places, e.g., alicantensis, strasburgensis, parisensis, utrechtensis, yorkensis, regensburgensis (MacAdoo, 1993).

Many localities (mostly lakes, rivers, seas, valleys, islands, capes, rocks or mountains, but also some towns or cities) have names that consist of two words, usually an adjective and a substantive (noun), e.g., Deep Lake, Black Sea, Dead Sea, Red River, Rio Grande, Rio Tinto, Long Island, Blue Mountain, Baton Rouge etc., or of two substantives, e.g., Death Valley, Lake Windermere, Loch Ness, Martha's Vineyard, Ayers Rock, Woods Hole, Cape Cod etc. Although such epithets would be correct in the sense of the Bacteriological Code, formation of specific epithets from such localities' names may pose a problem, because the use of the adjectival suffix -ensis, -ense may lead to rather strange looking or awkward constructions, such as “deeplakensis” or “bluemountainense”. If the name of a locality lends itself to translation into Latin, specific epithets may alternatively well be formed as genitive substantives by forming the genitives of the two components and concatenating them without hyphenation, e.g., like the existing ones lacusprofundi (of Deep Lake), marisnigri (of the Black Sea), marismortui (of the Dead Sea), or (of two nouns) vallismortis (of Death Valley). Note that in Latin the basic noun comes first, the determining word (adjective or noun) second. If possible one should avoid the inclusion of articles such as the, el, il, le, la, de, den, het, der, die, das, or their plurals los, les, las, ils, gli, le, de, die, etc. as they are used for locations in several languages, e.g., La Jolla, La Paz, El Ferrol, El Alamein, Le Havre, The Netherlands, Die Schweiz, Den Haag, Los Angeles, etc. Articles would unnecessarily elongate names without adding information.

F. Formation of names for bacteria living in association or symbiosis with other biota  An enormous reservoir of bacteria for future research is the microflora that is more or less tightly associated with other biota. I predict that at least two million new species (Trüper, 1992) will be described for the gut flora of various animal species.

Also the plant microfloras have so far been mainly investigated with respect to nitrogen fixation and diseases of economically important plants. To date, little has been done to investigate the phytopathogens that attack economically unimportant plants or weeds.

It is to be expected that microbiologists working in these fields will want to give new isolates names that relate to their hosts or associates.i.e., Latin nomenclatural names of animals, fungi, plants, and protists have been, and to a much larger extent, will be used.

This area of bacterial name-giving is unfortunately full of traps. Clearly, naming a bacterium after a host animal bearing a tautonym (such as Picus picus, the woodpecker) is easier than having to choose between generic name and a different specific epithet of the host. It is therefore important to know what these mean and how they were formed (adjective, substantive in genitive, etc.), in order to avoid nasty, ridiculous, or embarrassing mistakes.

The following example may demonstrate this situation: Certainly a bacterium isolated from the common house fly Musca domestica should not receive the epithet domesticus, -a, -um (“pertaining to the house”); its epithet should rather be muscae (of the fly) or muscicola (dwelling in/on the fly) the latter being a nominative noun in apposition.

The domestica associated with Musca is an adjective. If we theoretically consider it an independent noun meaning “the one pertaining to the house” one could, of course, form the genitive from it and thus produce a bacterial epithet domesticae. In this example, however, that would not make much sense as too many things “pertain to a house”. But formally it would not violate the Rules of the Bacteriological Code.

The easiest way of forming such specific epithets is the use of the genitive case of the generic name of the eucaryote in question, e.g., suis, equi, bovis, muscae, muris, aquilae, falconis, gypis, elephantis (of the pig, horse, cow, fly, mouse, eagle, falcon, vulture, elephant), or: fagi, quercus (fourth declension genitive, spoken with long u), castaneae, aesculi, rosae, liliae (of the beech, the oak, chestnut, horse chestnut, rose, lily).

Be aware of irregular forms such as bos (the cow), genitive: bovis, plural genitive boum! Use dictionaries and look up the declension in MacAdoo 1993!

G. Names taken from languages other than Latin or Greek  Besides names of persons or localities, many words from languages other than Latin or Greek have been used in bacterial names and certainly will be in the future. Here a few examples may suffice to demonstrate the width and variety of such cases:

During late medieval and renaissance times alchemy became rather fashionable among European scientists and many Arabic words entered into the terminology that would eventually be used in chemistry. One of these, which is often used in bacterial names, is “alkali” (Arabic al-qaliy, the ashes of saltwort) from which the element kalium (K, English: potassium) received its name. As the -i at the end of the word belongs to the stem it is wrong to speak and write of alcalophilic instead of alkaliphilic microbes. Latinized names of bacteria containing this stem should therefore be corrected to, e.g., Alkaligenes, alkaliphilus, etc., and new ones should be formed correctly!

A rather common mistake occurs with the English suffix - philic (e.g., hydrophilic—friendly to water, water-loving). This is clearly an English transformation of the Latin -philus, -a, -um (originating from Greek philos, friendly). All names formed thus far ending on -philicus, -a, -um are wrong and should, in my opinion, be changed to -philus, -a, -um as soon as possible. Here, however, Rule 57a (accordance with the rules of Latin) would have to be weighed against Rule 61 (retaining the original spelling) of the Bacteriological Code.

National foods or fermentation products often do not have equivalent Latin names and if typical microorganisms found in them or causing their fermentations are described, they have been (and may be) named after them, e.g., sake, tofu, miso, yogurt, kvas, kefir, pombe, pulque, aiva, etc. However, these names cannot be used unaltered as specific epithets in the form of nominative substantives in apposition (Trüper and De' Clari, 1997). They must be properly Latinized. The best way to do so is to form a neuter substantive from them by adding -um (e.g., sakeum, tofuum, kefirum, pombeum, etc.) and use the genitive of that (ending: -i) in the specific epithet (e.g., sakei, tofui, kefiri, pombei, etc.)

Another point worth mentioning is the “unnecessary” usage of words from languages other than Greek or Latin. For instance, the formation of the epithet simbae from the East African Swahili word simba, lion, for a Mycoplasma species was not necessary because in this genus the corresponding Latin epithet leonis (of the lion) had not been used before.

H. Formation of bacterial names from names of elements and compounds used in chemistry and pharmacy  The almost unlimited biochemical capacities of bacteria is another rather inexhaustible source for new names. Many generic names, as well as specific epithets, have been formed from names of chemical elements, compounds and even pharmaceutical and chemical products or their registered or unregistered trade names.

The late Robert E. Buchanan (1960, reprinted 1994) listed numerous examples of such generic names and specific epithets. Based on the classical Latin/Greek thesaurus and enriched by numerous Arabic words, the pharmaceutical sciences have, since the Middle Ages, developed a Neo-Latin terminology for chemicals of all categories.

The second largest category of chemicals are treated as neuter nouns of the third declension: these end on -ol (the alcohols), -al (aldehydes), -er (ethers, esters), and -yl (organic radicals); latinization does not change their names at the end, whereas the genitive is formed by adding -is.

Anions ending in -ite and -ate are treated as masculine nouns of the third declension. The English ending -ite is latinized to -is, with the genitive -itis, e.g., nitrite becomes nitris, nitritis. The English ending -ate is latinized to -as, with the genitive -atis, e.g., nitrate becomes nitras, nitratis.

Only few chemicals have names that are latinized in the first declension as feminine nouns, ending on -a with a genitive on -ae. Besides chemicals that always had names ending on -a (like urea), these are drugs found in classical and medieval Latin, such as gentian (gentiana) and camphor (camphora), further modern drugs, whose Latin names were formed by adding -a, like the French ergot becoming ergota in Latin.

The most important group of this category are alkaloids and other organic bases, such as nucleic acid bases and amino acids with English names on -ine. In Neo-Latin this ending is -ina, with the genitive -inae, e.g., betaina, -ae; atropina, -ae; adenina, -ae; alanina, -ae; etc.

For their use in bacterial generic names and specific epithets word stems and genitives of latinized chemical names are the basis. In principle they are then treated like any other word elements.

I. Arbitrary names  Either genus names or specific epithets “may be taken from any source and may even be composed in an arbitrary manner” (Bacteriological Code, Rule 10a and Rule 12c). They must, however, be treated as Latin. These “rubber” paragraphs open up a box of unlimited possibilities for people whose Latin is exhausted. But in view of the million names that will have to be formed in the future they are a simple necessity, whether Latin purists like them or not.

Examples for arbitrary generic names are Cedecea, Afipia, and in the near future “Vipia” and “Desemzia”, that were derived from the abbreviations CDC (Center for Disease Control), AFIP, VPI (Virginia Polytechnical Institute), and DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen), respectively. Examples for arbitrary specific epithets are, e.g., (Salmonella) etousae, derived from the abbreviation ETOUSA (European Theater of Operations of the U.S. Army), and (Bacteroides) thetaiotaomicron, formed from the three Greek letter names theta, iota, and omicron.

More recently, the new genus Simkania was described. The name is a latinized contraction of the first and the family name of the microbiologist Simona Kahane. Certainly an arbitrary name, short, elegant and easy to pronounce, points to future possibilities of bacterial name-giving. Authors should aim at such easily spelled and pronounced short names, when they take advantage of arbitrary name-giving.

III. Some case histories of malformed names

From the viewpoint of classical Latin many of the existing bacterial names are, plainly said, lousy in their grammar and etymology. However, under the Rules of the Bacteriological Code they are acceptable. A few case histories of wrong bacterial names are worth mentioning in a chapter on etymology because of their scurrility.

Acetobacter xylinus: This specific epithet goes back to Brown 1886, who described a Bacterium xylinum. Several subsequent changes of the genus (Trevisan 1889, Bacillus xylinus; Ludwig 1898, Acetobacterium xylinum; Pribram 1933, Ulvina xylina) prove by the change in gender that the epithet is an adjective. Because before 1951 (Bacteriological Code, Opinion 3), the gender of names ending in -bacter was not fixed as masculine, Acetobacter xylinum (Holland 1920 and Bergey et al. 1925) (all names and dates before 1950 cited were taken from Index Bergeyana, Buchanan et al., 1966) was not wrong either. As a consequence of Opinion 3 the species should be named Acetobacter xylinus. The Approved Lists of names (Skerman et al., 1980), however, listed the organism as Acetobacter aceti subspecies xylinum! Yamada 1983 revived the species status and correctly called it A. xylinus. The compiler of Validation List 14 (International Journal of Systematic Bacteriology, 1984) incorrectly put a sic after xylinus and changed it to the neuter form xylinum! (The Latin expression sic is used to point out a mistake or other peculiarity.) Unexpectedly the previous authors obeyed this falsifying change and even tried to give the neuter epithet justification by explaining it as a nominative noun in apposition (xylum, M.L. neut.n. cotton). “Acetobacter, called the cotton” makes little sense and certainly does not meet the requirements of a nominative noun in apposition (cf. Trüper and De' Clari, 1997), Finally, Euzéby 1997 corrected the name to A. xylinus.

Methanobrevibacter arboriphilus: In 1975 the new species Methanobacterium arbophilicum was described. The organism was isolated from rotting trees and the authors wanted to express “friendly to trees” by the epithet. In Latin, tree is arbor, genitive arboris, ie., the stem is clearly arbor-, not arbo-. The second error was that the English ending -philic was latinized to -philicum instead of correctly to -philum. Although this was first pointed out to the authors in 1976, they did not correct the epithet themselves. Then, in a review paper, Balch et al. 1979 rearranged the methanogenic procaryotes and transferred the species to the genus Methanobrevibacter as M. arboriphilus (the correct form of the epithet). It was again the compiler of the Validation List No. 6 (International Journal of Systematic Bacteriology, 1981), who created a new wrong form of the epithet, arboriphilicus! Although immediately informed of his error, he did not correct it. And so this wrong epithet still occurred in Bergey's Manual of Systematic Bacteriology, Vol. 3 (1989). To my knowledge it has not been corrected!1

Some time ago an author wanted to create the specific epithet “nakupumuans” and explained this word as derived from the Maori word nakupumua, breaking protein down to fragments.

Becoming informed that there was neither need to use another language than Latin, nor any specific connection between the Maori and protein degradation the author decided to call the isolate proteoclasticum. Accepting such name formations in procaryote nomenclature would mean giving up Latin as the basic language of biological nomenclature. As long as names can be formed from the Latin/Greek thesaurus at our hands, names from other languages should be avoided.

In another instance, an author wanted to propose a specific epithet in honor of a colleague and formed an epithet ending in -icus. As this is not within the Rules, I advised him to choose either an epithet ending on -ii (genitive noun) or on -ianus (adjective). His answer was that he did not like the former and felt that the latter sounded like an insult to the colleague to be honored!

Another colleague correctly formed the generic name Acidianus (accentuation: a.cid.ia'nus) from the Latin neuter noun acidum, acid and the Latin masculine noun Ianus, the Roman god with the two faces, by which he wanted to indicate the ability of the organism to both oxidize and reduce elemental sulfur. With this spelling the epithet promptly became mispronounced (a.ci.di.a' nus) suggesting a different meaning and causing suggestive jokes. Here the use of the consonantic i, (i.e., j) would have sufficed to suppress the misinterpretation: Acidijanus would be the choice.

These examples also show that nobody is free from making mistakes. During my work in this field I have made several, and sometimes even given wrong advice, quite to my embarrassment afterwards.

IV. Practical etymology in descriptions of genera and species

As mentioned before, for the average microbiologist “etymology” is a kind of nasty linguistic exercise necessary for the description of a new genus or species. In reality he/she has to “create” a new name; the organism has been isolated and determined by the author, not “created”! The better and more modest wording would be, to “propose” a new name.

The abbreviation M.L. is very often misunderstood as medieval Latin. I personally would therefore prefer a ruling that M.L. would really mean medieval Latin and that modern Latin, better Neo-Latin, would be abbreviated N.L.

V. Recommendations (from the viewpoint of language) for future emendations of the Bacteriological Code

We should not aim for pure classical Latin in biological nomenclature but rather develop the current Latin/Greek thesaurus further by following the Rules of the ICNB or the respective codes of nomenclature applicable to other fields of biology. This is in reality what has happened since Linnaeus' time. In my opinion the ICNB has excellent provisions to do so. This is already documented by the low number of Opinions that had to be issued by the Judicial Commission of the ICSB during the last ten years.

For several years the development of a uniform code of nomenclature for all biological taxa has been underway, enlisting the participation of well-known taxonomists from bacteriology, botany, mycology, phycology, protozoology, virology, and zoology. This effort has received the support of the International Unions of Biological and Microbiological societies, IUBS and IUMS (Hawksworth and McNeill, 1998). These activities reflect the general scientific need to assess the total extent of biodiversity on Earth, in order to facilitate conservation and, perhaps, prevent further extinction of the biota. For this purpose a unified system of biological names has been considered indispensable. Drafts of the future universal “BioCode” have been published, the latest (fourth) draft by Greuter et al. 1998. As soon as the BioCode is accepted by the taxonomic committees of the different biological disciplines involved, the Bacteriological Code will have to be revised to conform with any new recommendation. Changes in etymological rulings should be expected. Unfortunately the recommendations for latinization (Articles 37–39) are not yet formalized, therefore comments and recommendations cannot be offered at this time.