Shakespeare once sagely wrote, ‘What's in a name? That which we call a rose by any other name would smell as sweet,’ but this fine philosophical spirit is not always appreciated by students and amateur botanists faced with changes of name of familiar native plants. Most people who are interested enough in plants to want to recognise and name the different species realise that the system of Latin names is essential for accurate scientific study, but are discouraged by confusing name changes which they do not understand. Taxonomists, the botanists who classify and name plants, are often regarded as taking delight in ‘lumping’ and ‘splitting’ species and making unnecessary name changes, but this is a grossly unfair picture. The rules they have devised for naming plants aim above all at stabilising names and removing ambiguities and confusions, and even a superficial acquaintance with these rules will show what a difficult task taxonomists have to determine the one and one only correct name for every plant. Changes in established names are made reluctantly and only when absolutely necessary; to a large extent the botanists of the present must suffer in this way for the confusion and errors of the past when there were no rules. If the logical reasons behind the changes are understood they should not be resented.

Before name changes can be explained the structure of the names must be properly understood. The Latin names of the convenient natural units we term species are binomials, that is they are composed of the name of the genus or broad group to which the species belongs followed by the specific epithet which names the species and is peculiar to it. This is roughly comparable to putting one's surname first and Christian name second, and just as the same Christian names are used hundreds of times in different families so specific epithets can be used again and again in different genera, but names of genera can only be used once. The names of genera are nouns, and specific epithets are usually adjectives or possessive nouns. If they are adjectives, as most are, they must agree in gender with the name of the genus they are combined with. The name of the silver tree-fern, for instance, is Cyathea dealbata, Cyathea being the name of a genus of tree-ferns and dealbata the descriptive specific epithet, agreeing in feminine gender with Cyathea. It is easier to remember specific epithets if their meaning is understood. They usually describe some outstanding feature of the plant or its habitat, or commemorate its discoverer or the locality where page 85 it grows, or its resemblance to another plant. The epithet dealbata, for instance, literally means ‘whitish’, and refers to the silvery-coloured undersides of the leaves in the silver tree-fern. The Botanical Names of the Flora of New Zealand by Wall and Allan is a useful little book giving full information about the meaning and origin of generic and specific names. Taking specific epithets literally can be misleading if they are based on a feature which is not constant or is shared by several species. The botanist who first named our native taupata met it on exposed coastal rocks where it is completely prostrate and so named it Coprosma repens, or the creeping Coprosma. In more sheltered situations it grows into a tree-like shrub up to 15 feet high, so that its name in these cases is most inappropriate. The first eye-bright named in New Zealand was called Euphrasia cuneata because of its cuneate or wedge-shaped leaves. As most of the other New Zealand species of Euphrasia also have leaves which are cuneate to some extent this epithet is not particularly helpful.

Until recently botanists used to distinguish specific epithets based on proper names by giving them a capital initial letter, such as Olearia Colensoi for the daisy tree named after William Colenso. For the sake of uniformity the present practice is to use a small letter for all specific epithets. The name of a genus always has a capital letter, and it is permissible to abbreviate a generic name to its capital letter after the first use unless it could possibly be confused with another genus under discussion beginning with the same letter. It is customary to italicise Latin names in print, but names of genera used as common names are put in ordinary type without a capital. For instance, Coprosma repens is the name of a native plant but it is correct to talk of a coprosma hedge.

A binomial plant name is not fully cited unless it is followed by the name of the authority, that is the botanist who gave the specific epithet. This is usually an abbreviation, as in Coprosma repens A.Rich. where A.Rich. stands for Achille Richard, an early French botanist who worked with New Zealand plants. A botanist who described numerous species of New Zealand plants was Sir Joseph Hooker. His name is abbreviated as Hook.f., standing for Hooker filius (Hooker the son), to distinguish him from his father Sir William Hooker. When a species originally described under one genus is transferred to another genus the name of the author of the epithet is placed in brackets followed by the name of the botanist transferring the species. The earliest name of the pohuehue is Polygonum complexum A.Cunn. When Meissner decided that it ought to be taken out of the genus Polygonum and placed in the related genus Muehlenbeckia he made the new combination Muehlenbeckia complexa (A.Cunn.) Meissn. (note the change in gender of the specific epithet).

Though ‘species’ is in general the most useful category, categories of lower rank can be used to distinguish groups of less importance within a species. The most common infraspecific category is ‘variety’, abbreviated to ‘var.’, and varieties are named with epithets in the same way as species. Just as it is meaningless to quote a specific epithet without the name of the page 86 genus it is placed under, a varietal epithet must be associated with the name of the species to which it belongs, so forming a trinomial. For instance, if a botanist finds prostrate plants of a normally erect species he may group them together as a variety under the epithet prostrata, just as prostrate plants of Epilobium microphyllum have been named E. microphyllum var. prostratum Petrie. The species is then considered to consist of two varieties, an erect one and a prostrate one. Varieties can come to be regarded as species and vice versa, and this is known as a change of rank. When the rank of a species or variety is changed, the original epithet being retained, the name of the author of the epithet is again placed in brackets followed by the name of the botanist making the change. The big five-finger was originally described as Nothopanax arboreum var. laetum Kirk. When Cheeseman decided it differed enough from ordinary Nothopanax arboreum to be a separate species he made the new combination Nothopanax laetum (Kirk) Cheesem.

The binomial system of nomenclature we are so familiar with today was first seriously used by the Swedish botanist Linnaeus in the middle of the eighteenth century, and was such an improvement on the haphazard systems previously used that it was quickly adopted and developed by other botanists. Though Linnaeus published his own nomenclatural principles there were no generally accepted rules at this stage, and as more and more plants were named confusion of names began to occur. On the whole botanists sensibly tried to avoid giving the same epithet to two different species in the one genus, and if they discovered that one plant had been given two names the earliest one was usually recognised as having the best claim, but distribution of literature was very slow by today's standards and it was not surprising that they constantly duplicated each other's names and named the same plant several times. At the first International Botanical Congress in 1867 an attempt was made to draw up an international code of rules for naming plants, but botanists disagreed in their opinions and for more than the next half-century different rules were applied in different countries. The differences were gradually ironed out and by the time of the fifth I.B.C. in 1930 agreement had been reached on all essential points, so that the third edition of the International Rules of Botanical Nomenclature then produced was truly internationally accepted and applied. The rules were revised by successive Congresses and the most recent edition is the ‘Paris Code’ adopted by the eighth I.B.C. at Paris in 1954 and published in 1956. Subsequent editions will be revised in their turn, for provision for modification is an important feature of the Code. Except in special cases the rules set out in the Code are retroactive, and are used to put past nomenclature in order as well as providing for future work.

In all matters of nomenclature, therefore, the current edition of the International Code of Botanical Nomenclature (as it is now titled) is the botanist's bible, and it is officially printed in English, German and French, and also translated into Spanish and Russian. Its precise phraseology and intricate provisions for every contingency are rather difficult for anyone but page 87 professional taxonomists to understand, so some of the most relevant and important portions are quoted and explained below. Throughout the Code the rules and recommendations are illustrated by examples, mostly using Northern Hemisphere plants. These examples have been replaced below with similar ones from our own flora to show how the rules affect us in practice.

The first Division of the Code contains the six Principles which are the basis of the system of botanical nomenclature, while the second and largest Division deals with the Rules by which the Principles are applied. The three Principles most relevant to this article are:

II.	The application of names of taxonomic groups is determined by means of nomenclatural types.
III.	The naming of taxonomic groups is based on priority of publication.
IV.	Each taxonomic group can bear only one correct name, the earliest that is in accordance with the Rules, except in specified cases.

(The first article of the Rules explains that throughout the Code the word ‘taxon’, plural ‘taxa’, will be used to denote taxonomic groups of any rank. Full rules for taxa of all ranks are provided but only those referring to species and varieties will be discussed here.)

Types. The type system as put forward in Principle II is the essential link tying the system of nomenclature to the actual plants. Article 7 defines a nomenclatural type as that constituent element of a taxon to which the name of the taxon is permanently attached.

A note explains that this is not necessarily the most typical or representative element of a taxon. The type of Carmichaelia williamsii Kirk, for instance, is a specimen in the Kirk Herbarium at the Dominion Museum, Wellington, which was sent to Kirk by Bishop Williams. The type specimens of the many species described by Sir Joseph Hooker were sent to him at the Royal Botanic Gardens, Kew, in England and are still in the Herbarium there. The Code defines several different kinds of types.

A holotype (‘type’) is the one specimen or other element used by the author or designated by him as the nomenclatural type. For so long as a holotype is extant, it automatically fixes the application of the name concerned.

When no holotype was chosen by the author or it is lost or destroyed a substitute for it is chosen which may be either a lectotype or a neotype, a lectotype being chosen from the original material and a neotype from other material for so long as all the original material is missing. (Fuller notes on type nomenclature as applicable in Zoology have been given in Tuatara Vol. IV no. 2.) Type specimens are carefully preserved in the large herbaria of the world and are available for reference. Any taxonomist working with a group of species must study the type specimens to which the names are attached, or choose types if there are none, before he can come to any conclusion about the application of the names. Though every specimen placed in a certain species need not match the type in every detail, whatever limits are decided upon for the species must include the type.

page 88

When a species is divided into two or more varieties one of the varieties must include the type of the species and have this for its own type. It is known as the ‘type variety’ and must bear the same epithet as the species. This rule is stated in Article 25:

For nomenclatural purposes, a species or any taxon below the rank of species is regarded as the sum of its subordinate taxa if any. Valid publication of a subordinate taxon which does not include the nomenclatural type of the higher taxon automatically circumscribes a second taxon of the same rank which has as its nomenclatural type the type of the higher taxon …and bears the same epithet.

For example, Geranium traversii Hook.f. is a normally white-flowered species from the Chatham Islands which has as type specimen a white-flowered Travers specimen at Kew. When Cockayne described pink-flowered variants of this species as var. elegans he at the same time automatically created var. traversii containing all the plants matching the type in having white flowers. The type of var. traversii is the same Travers specimen which is the type of the whole species.

Choosing and Changing Names. Principles III and IV are the basis of the rules dealing with choice, retention and rejection of names. A large proportion of the name changes suffered by many of our plants are merely the replacement of previously used incorrect names with the correct one according to these rules. The limits of the taxa remain the same, and the changes are of name only in the strictest sense. Article 11 formally defines a correct name and states the rule of priority as it applies to species and taxa of lower ranks:

For any taxon below the rank of genus, the correct name is the combination of the earliest available legitimate epithet validly published in the same rank with the correct name of the genus, species, or taxon of lower rank to which it is assigned.

A ‘legitimate’ name is defined as one that is in accordance with the rules, while an ‘illegitimate’ name is contrary to the rules. What is termed ‘valid publication’ is clearly defined in another section and Article 12 states:

A name of a taxon has no status under this Code unless it is validly published.

That is, unless a name was published validly as described in the Code it cannot be considered in the system of priority at all or even classed as legitimate or illegitimate but is completely ignored. This is to ensure that names are published with proper descriptions in publications available to botanists so that there is no doubt what they are meant to apply to. The name of a taxon published without an accompanying description is known as a nomen nudum. No name published before 1753, the date of Linnaeus’ Species Plantarum, has any standing under the Code, this date having been fixed arbitrarily as the starting point for the system of priority.

A taxonomist in search of the correct name of a species therefore first finds all the validly published names which have been given to the species and, if possible, examines all the type specimens supporting the names to be sure that they really are all included within what he considers are the limits page 89 of the species. He then sets the names in chronological order and, starting from the earliest, investigates whether they are legitimate or illegitimate. He must keep discarding the illegitimate names until he comes to the earliest legitimate one, or if all the names given to a species are illegitimate he must provide a new one. According to Article 62 he cannot discard a legitimate epithet

merely because it is inappropriate or disagreeable, or because another is preferable or better known, or because it has lost its original meaning.

The earliest validly published epithet may be illegitimate for a number of reasons, most commonly

If it is a later homonym, that is if it duplicates a name previously and validly published for a taxon of the same rank based on a different type. Even if the earlier homonym is illegitimate, or is generally treated as a synonym on taxonomic grounds, the later homonym must be rejected. (Article 64.)

To find out if there is an earlier homonym for any name Index Kewensis must be consulted. This is a huge list of all the names of genera and species of flowering plants ever published which is brought up to date with a supplement every five years. There is no comparable list for ferns. Examples of names of New Zealand plants being abandoned because of the discovery that the epithet had been used before in the same genus are Senecio robustus Buchanan (1874) being renamed S. revolutus by Kirk because of S. robustus Sch.Bip. (1845), an Asian species, and, more recently, Epilobium parviflorum Simpson and Thomson (1943) being renamed E. pratense Simpson because of E. parviflorum Schreb. (1771). Accidental duplication of epithets like this is most likely to occur in genera such as Senecio and Epilobium which have a wide range in several continents. More unexpectedly a recently described New Zealand orchid Corybas saprophyticus Hatch (1952) had to be provided with a new name (C. cryptanthus) because of C. saprophyticus Schlecht (1924). With the literature now available mistakes of this kind should no longer occur. Provided the binomials were fully cited with their authorities, the previous uses of homonyms in literature should not be confusing, even when the identically named species come from the same region.

If the earliest epithet of a species is illegitimate (or has to be discarded for other reasons given in the rules) the later epithets (if any) must be considered in their turn until the earliest legitimate one is found. Names published after the earliest legitimate one (provided they were applied to exactly the same species) are automatically illegitimate because they were unnecessary. This is stated formally in Article 64: a name is illegitimate

If it was nomenclaturally superfluous when published, i.e. if the taxon to which it was applied, as circumscribed by its author, included the type of a name or epithet which ought to have been adopted under the rules.

Widely used names of many of our native plants are incorrect because they are illegitimate later synonyms, and name changes purely on the grounds of priority are common and regrettable but should become less frequent as the early literature relating to our flora is thoroughly examined.

page 90

Though it is incorrect to continue using the familiar synonyms now that they are known to be illegitimate, their use in past literature does not cause confusion, as in each case both the earlier and later names refer only to the one species. Thus though the pohutukawa has long been known as Metrosideros tomentosa A.Rich. (1832) its earliest and therefore correct name has been found to be M. excelsa Sol. ex Gaertn. (1788). Similarly the name Clematis indivisa Willd. (1800) for our beautiful native puawhananga must give way to the earlier C. paniculata Gmel. (1791).

Authors often provided new illegitimate epithets instead of using the earliest when transferring a species from one genus to another. Previously it was considered allowable or even correct in these cases to provide a new epithet for the species in the new genus, but Article 55 of the present Code states:

When a species is transferred to another genus (or placed under another generic name for the same genus) without change of rank, the specific epithet, if legitimate, must be retained, or (if it has not been retained) must be reinstated, unless one of the following obstacles exists: (1) that the resulting binary name is a later homonym … (2) that there is available an earlier legitimate specific epithet.

The earliest legitimate epithets of many of our species were overlooked or ignored for many years because they were under different genera from those in which they are usually placed. Thus the heketara was named Olearia cunninghamii by Hooker in 1864 but it had already been named Brachyglottis rani by Cunningham in 1838 and the combination Olearia rani (A.Cunn.) Druce is the correct name. Coprosma grandifolia Hook.f. (1853) has had to give way to C. australis (A.Rich.) Robinson based on Ronabea australis A.Rich. (1832). Hooker was incorrect in naming the wineberry Aristotelia racemosa (A.Cunn.) Hook.f. (based on Friesia racemosa A.Cunn. (1839)) when the earliest name was Dicera serrata J.R. et G.Forst. published in 1776. A. serrata (J.R. et G.Forst.) Oliver is the correct name. As foreseen in Article 55 sometimes the earliest legitimate epithet given under one genus cannot be used when the species is transferred to another because it would become a later homonym. In 1882 Buchanan mistakenly named a plant belonging to the Veronica complex Mitrasacme cheesemanii. When Kirk wished to transfer the species to Veronica in 1896 he was forced to provide a new name (V. quadrifaria) because of V. cheesemanii Benth. (1881), quite a different species. Even though Bentham's species was later transferred to Parahebe the combination V. cheesemanii cannot be used again and V. quadrifaria is the correct name for Buchanan's plant as long as it remains in Veronica. When it was transferred to the genus Hebe by Cockayne and Allan in 1926 the earliest epithet was again available and the correct combination made was Hebe cheesemanii (Buchan.) Ckn. et Allan.

An example from the genus Coprosma shows how proper application of the rules concerning priority and later homonyms to a nomenclatural tangle leads to three accepted names having to be discarded. The taupata was first named Coprosma repens by Richard in 1832. Hooker, not realising page 91 it was the plant already named by Richard, illegitimately renamed it C. retusa in 1844, but subsequently decided that it was after all identical with a Norfolk Island species C. baueri Endl. (1841) (which he mistakenly rendered C. baueriana) and for the next seventy years or so it was commonly known by one of the last two names. Then Oliver decided that the New Zealand plant was different from the Norfolk Island C. baueri and, after investigating the type of Richard's C. repens, correctly revived repens as the earliest legitimate epithet. Now in 1844 Hooker had illegitimately re-used the epithet repens for a creeping alpine Coprosma. When the earlier use was discovered C. repens Hook.f. had to be discarded and the alpine species is correctly known by the later name C. pumila Hook.f. Likewise the epithet retusa was used again by Petrie in 1894 for a species quite different from taupata. Although C. retusa Hook.f. is an illegitimate later synonym of C. repens A.Rich. the combination cannot be used again and C. retusa Petrie must be replaced by the name C. crenulata provided by Oliver.

There is no room in standard taxonomic works for such long nomen-clatural explanations as are given in the above paragraph. Instead taxonomists use a sort of shorthand which, if properly understood, conveys just as much information. The exact style varies but it is usual to list the name considered correct first, followed by the synonyms. The nomenclatural history of taupata can be summarised as set out below, for brevity the date of publication only being given instead of the full literature reference usually included:

• Coprosma repens A.Rich. (1832) non Hook.f. (1844)

  • C. retusa Hook.f. (1844) non Petrie (1894)

  • C. baueri auct. non Endl. (1841) (quoted in error by Hooker as baueriana).

(The abbreviations ‘auct.’ (or ‘auctt.‘) and ‘auth.’, standing for auctores and authors, indicate that an epithet has been used in a particular sense by a number of authors.)

Though illegitimate names must be rejected, an epithet originally published as part of an illegitimate name may be made legitimate later in another combination. This is stated in a note to Article 72:

When a new epithet is required, an author may, if he wishes, adopt an epithet previously given to the taxon in an illegitimate name, if there is no obstacle to its employment in the new position or sense; the epithet in the resultant combination is treated as new.

For example, Kirk in 1877 named a small creeping native Veronica V. canescens, but this name was illegitimate as the combination V. canescens had been used previously three times in different senses. The species was renamed V. lilliputiana by Stearn in 1951 and lilliputiana remains the earliest legitimate epithet under the genus Veronica. However in 1944 Oliver had transferred the species to the genus Parahebe and the binomial P. canescens Oliver, treated as a new name dating from 1944, is legitimate though the combination P. canescens (Kirk) Oliver dating back to 1877 page 92 is not. The earliest legitimate epithet under the genus Parahebe is therefore canescens. If the epithet lilliputiana had been bestowed before 1944, however, it would have taken precedence in Parahebe also. A similar case with the events in the reverse order occurred in Metrosideros. Melaleuca lucida Forst.f. (1786), the earliest name of the southern rata, is an illegitimate later homonym of Melaleuca lucida Linn.f. (1781), a Pacific plant. The New Zealand species was transferred to Metrosideros by Richard in 1832 and the illegitimate combination Metrosideros lucida (Forst.f.) A.Rich. was used for a long period. The binomial Metrosideros lucida A.Rich. treated as a new name dating from 1832 could, in fact, have been legitimate except that there was an earlier name Metrosideros umbellata Cav. (1795) which takes precedence as the correct name of the species.

Varietal epithets rarely have to be discarded because they are later homonyms; it is legitimate (though not recommended) to use the same epithet for varieties of different species in the same genus. It is also unusual for a variety of one species to be given two different epithets, However, application of Article 27 which states,

In the name of an infraspecific taxon which includes the nomenclatural type of the epithet of the next higher taxon, the epithet of this higher taxon must be repeated unaltered …without citation of an author's name. …

will cause the names of many varieties to be changed as this rule is comparatively recent. In the past frequently only the variety differing from the type would be named and described, the ‘type variety’ usually being described in the species description. If the type variety was described in parallel with the other varieties it was often given epithets such as ’ vera, typica, genuina’, which are not acceptable under the present Code. When Oliver divided Coprosma propinqua into two varieties he named one var. typica and the other var. martinii. But as var. typica includes the type of the species it must be called var. propinqua.

Changes of Taxonomic Status. The examples of name changes that have been discussed so far are merely the substitution of correct names for incorrect ones according to the rules and must be accepted as absolute as long as the rules have been correctly interpreted. Name changes of another kind occur when two species are merged or a species is split into two, or when a variety is raised to specific rank or a species placed as a variety of another, or, more formally, when the limits of taxa or their relations to other taxa are altered. Frequently all the names involved are legitimate. Changes of this kind cannot be considered absolute as they depend to a certain extent on the personal opinion of different botanists and may be reversed more than once. The amateur botanist usually has little chance of deciding for himself which of two conflicting interpretations is nearer the truth and is best advised to follow the usage in the current standard Flora. The complex sets of alternative names in ferns give many good examples of this. Ferns are a world-wide yet comparatively small group which have page 93 attracted the attention of a great number of botanists who often differ in their opinions as to the limits and therefore the names of taxa. For instance, the common fern often known as Blechnum procerum (Forst.f.) Swartz can also be considered merely a form of the African species B. capense (L.) Schlecht. Likewise bracken can be referred to as Pteridum aquilinum var. esculentum (Forst.f.) Kuhn or P. esculentum (Forst.f.) Diels. The change of name of the Kermadec pohutukawa, unlike most of the other recent name changes in the genus Metrosideros, is also based on a taxonomic decision. Previously it was considered merely a form of the common Pacific species M. villosa and appears under that name in Cheeseman's Manual. After studying the group Oliver decided it was really a different species and named it M. kermadecensis. The common harebell Wahlenbergia albomarginata Hook.f. was for a long time merged with the Tasmanian species W. saxicola DC. but is now once again considered distinct. According to the latest study the plants which in Cheeseman's Manual are referred to Oreomyrrhis andicola Endl. as three varieties, var. colensoi, var. ramosa and var. rigida, have been separated from that South American species and stand as three related endemic species O. colensoi Hook.f., O. ramosa Hook.f. and O. rigida (Kirk) Allan. To a large extent this is merely a return to the original treatment.

Although alteration of taxonomic status is the basis of these changes there are rules of nomenclature which must be followed when the changes are made. If, when a change of rank or of specific limits is made, the name selected is not the correct one in accordance with these rules, then a nomenclatural change will be necessary later in addition. The relevant rules, with explanatory examples, are as follows.

Artile 53: When a species is divided into two or more species, the specific epithet must be retained for one of them, or (if it has not been retained), must be reinstated. When a particular specimen was originally designated as the type, the specific epithet must be retained for the species including that specimen. When no type was designated, a type must be chosen….

In 1844 Hooker named a species of Coprosma from the Auckland Islands C. cuneata. In his work on the whole New Zealand flora in 1853 he enlarged his idea of the species to include slightly different North and South Island plants as well as the original Auckland Islands specimens, and this interpretation was followed in Cheeseman's Manual. In 1928 Oliver decided that this broad version of C. cuneata, or C. cuneata sens. lat., should be divided into two species, one the original Auckland Islands plant including the type which retained the epithet cuneata and became C. cuneata Hook.f. in the narrow sense (sens, strict.), and the other the North and South Island plant which he described as a new species C. pseudocuneata.

Article 57: When two or more taxa of the same rank are united, the oldest legitimate epithet is retained…. The author who first unites taxa bearing names or epithets of the same date has the right to choose one of them, and his choice must be followed.

For instance, if a future taxonomist decides once more that the New Zealand page 94 taupata should be united with the related Norfolk Island plant the earliest epithet for the whole group would be C. repens A.Rich. (1832). Previously they were incorrectly united under the later name of the Norfolk Island species, C. baueri Endl. (1841).

Article 60: When the rank of a genus or infrageneric taxon is changed, the correct name or epithet is the earliest legitimate one available in the new rank. In no case does a name or epithet have priority outside its own rank.

Though it is thus not mandatory for authors to retain the existing epithet when changing the rank of a taxon, recommendations to Article 60 urge that this should be done whenever possible, and it very frequently is. When Hooker's two species Oreomyrrhis colensoi and O. ramosa were made varieties of O. andicola by Kirk he retained the same epithets and made them var. colensoi (Hook.f.) Kirk and var. ramosa (Hook.f.) Kirk, though he need not have done so. However, when their rank was changed back to that of species again, as mentioned above, there was no choice about using their original epithets in that rank. When O. andicola var. rigida Kirk, which was originally described as a variety, was raised to the rank of species parallel with the others, again the original epithet was retained to make it O. rigida (Kirk) Allan. Because the names have been kept the same these changes are not nearly so troublesome as they might have been. Some revisions resulting in changes of rank are not so easy to follow. The Spaniard Aciphylla colensoi Hook.f. is treated in Cheeseman's Manual as a wide-ranging species with two varieties, var. conspicua Kirk and var. maxima Kirk. The complex is now treated as three separate species. The old var. conspicua, which actually contained the type of the species and should have been var. colensoi, when treated as a species becomes A. colensoi sens. strict. Var. maxima Kirk has permissibly been renamed A. scott-thomsonii Ckn. et Allan and A. aurea Oliver is a newly described third species formerly included also in Cheeseman's wide treatment of A. colensoi. This old broad specific complex can be referred to as A. colensoi sens. lat. In some cases, especially when a variety is being raised to specific rank, a name change may be unavoidable because the varietal epithet would be a later homonym if used as a specific epithet. For instance, when Hooker wished to make his Viola cunninghamii var. gracilis a species he named it V. lyallii, undoubtedly because he knew there were several previous uses of the binomial V. gracilis.

Because of Article 57, when a taxonomist wishes to class two taxa previously regarded as separate species as two varieties of one species he must retain the older of the two specific epithets as the name of the combined species. When N. E. Brown decided that Wahlenbergia pygmaea Col. (1899) was only varietally distinct from W. albomarginata Hook.f. (1852), he placed the former as W. albomarginata var. pygmaea (Col.) N. E. Brown, at the same time automatically creating var. albomarginata. He could not have made W. albomarginata a variety of W. pygmaea.

Misapplication of Names. We have already dealt with changes of name caused either by replacement of incorrect names with correct ones or page 95 changes of taxonomic status. Sometimes what appear to be similar ‘name changes’ are merely corrections of generally accepted misconceptions as to which name belongs to which species. These errors can be detected only by examining the type specimens supporting all the names concerned. Misapplication of another botanist's names has unfortunately occurred many times in the New Zealand flora, often because the early descriptions were short and unsatisfactory and the type specimens far away. Mistakes of this kind are still being found, with resulting confusion when they are corrected. Oliver, working with the genus Coprosma, found by looking at the type specimen that the species commonly known as C. depressa Col. ex Hook.f. was quite different from the one originally described under that name which was passing under the later name C. ramulosa Petrie. Colenso's species therefore had two names while the one commonly known as depressa really had none. To right this Oliver discarded C. ramulosa as an illegitimate later synonym of C. depressa and named the unnamed species C. cheesemanii, but the confusion as to what an author means when referring to ‘C. depressa Col. ex Hook.f.’ is less easily cleared up. A worse example of this sort of tangle occurred in the genus Rubus. G. Forster in 1786 described a species of bush lawyer as Rubus australis. We shall call this species (a). Cunningham in 1830 described two more species (b) and (c) and named them respectively R. cissoides and R. schmidelioides. Later botanists misinterpreted Forster's poor description and used his epithet australis for species (b) while Cunningham's epithet schmidelioides was used for species (a), species (c) being treated as a variety of species (a). To complete the confusion the epithet cissoides was then applied to a fourth species (d) which is really R. squarrosus Fritsch. The names were widely used in these wrong senses and appeared thus in Cheeseman's Manual. In 1935 Allan discovered the errors by examining the type specimens and published a table showing the relation between the correct and incorrect usages. It is now very difficult to know in what senses the names are being used, and until time has caused the wrong uses to be forgotten authors must make their meaning clear by explanatory citations, e.g. for species (a): R. australis Forst.f. (not of Cheeseman Man. N.Z. Fl. 1925, 500) = R. schmidelioides auct. non A.Cunn. Misapplications are not always as clear as this and it is sometimes difficult to draw a line between a misapplication and a difference of opinion about specific limits.

Names of Genera. The rules concerning the names of genera have not been discussed but many of them are similar to those governing specific epithets, with obvious exceptions following on the difference in rank of the two categories. Most of the changes made in generic names in the New Zealand flora are the result of differing opinions of generic limits and, like the alternative names and changes of rank of species and varieties mentioned earlier, are not necessarily to be accepted. The kowhai is usually placed in section Edwardsia of the genus Sophora, but some botanists prefer to treat Edwardsia as an independent genus. The group of species originally described under the genus Myrsine have been regarded by other authors as page 96 belonging to the more restricted genera Suttonia and Rapanea. Again the ferns provide numerous examples. The filmy ferns placed in Cheeseman's Manual under the genus Hymenophyllum can be split up into groups and placed under a number of different genera such as Mecodium, Meringium, etc., leaving only a few in Hymenophyllum sens. strict. The common polypody P. diversifolium has at different times been placed under the genera Polypodium, Microsorium and Phymatodes. Changes of generic names which are simply corrections of nomenclature are rare, though in the orchids Corybas Salisb. (1805) has replaced Corysanthes R.Br. (1810) purely on the grounds of priority. Stability of generic names is aided by a provision of the Code which allows generic names to be permanently conserved against earlier homonyms or legitimate synonyms if it can be shown that changing them would upset well-established and very widely-used names. For instance, when the monotypic New Zealand genus Shawia, founded in 1776 to contain the akiraho S. paniculata, was merged with the later-published but much larger genus Olearia (1802) the name Olearia was conserved against Shawia (for as long as the two were united) because it had been widely used for a large number of species. Other conserved generic names (or nomina conservanda) in the New Zealand flora are Nertera, Pterostylis, Wahlenbergia, Parsonsia and Persoonia. Specific epithets cannot be conserved in this way.

Literature

ALLAN, H. H., 1935— Nomenclature of the New Zealand Species of Rubus, in Notes on New Zealand Floristic Botany No. 6. Trans. Roy. Soc. N.Z. 65 (3), 221-231.

CHEESEMAN, T. F., 1925— Manual of the New Zealand Flora. 2nd Edition, Govt. Printer, Wellington.

OLIVER, W. R. B., 1935— The Genus Coprosma. Bernice P. Bishop Museum Bull. 132. Honolulu.

ROLLO, L. J., 1951— Type Nomenclature. Tuatara 4 (2), 54-55.

WALL, A., and ALLAN, H. H., 1950— The Botanical Names of the Flora of New Zealand. 2nd Edition. Whitcombe and Tombs.

Index Kewensis Plantarum Phanerogamarum. 1995 and later supplements. Oxford.

International Code of Botanical Nomenclature (adopted by 8th Intern. Bot. Congr., Paris, 1954). 1956. Utrecht.

For Further Reading

LAWRENCE, G. H. M., 1951— Taxonomy of Vascular Plants. Chapter IX, Plant Nomenclature, 192-222. Macmillan, New York.