Cunningham’s Florae Insularum Novae Zelandiae Precursor and the correct author of the fern genus Loxsoma nom. cons.

ABSTRACT

Allan Cunningham’s Florae Insularum Novae Zelandiae Precursor, published serially from 1837 to 1839, is an important source of taxonomic names in New Zealand botany. It is pointed out that the printed text contains a large number of errors, which can be corrected using an autograph manuscript that is still extant. In particular, the spelling of the name of the fern genus Loxsoma nom. et orth. cons. as Loxoma is discussed. It is shown that this is an orthographic error, soon corrected by its author, who is not Cunningham, as generally supposed, but W.J. Hooker, as actually stated in the published text. The manuscript confirms the attributions Loxsoma R.Br. ex Hook. and L. cunninghamii R.Br. ex Hook. A number of typographic errors are also discussed, the only significant one having resulted in the mistaken attribution of the family Griseliniaceae to Cunningham.

KEYWORDS:

• Allan Cunningham
• Loxsoma
• New Zealand
• orthographia conservanda
• William Jackson Hooker

Introduction

The first published English language synopsis of the flora of New Zealand was Allan Cunningham’s Florae Insularum Novae Zelandiae Precursor (hereafter Precursor), which appeared in 17 parts in two different journals over a period of 2 years and 9 months, beginning in 1837 (Cunningham 1837–1839). As one of the earliest New Zealand Floras, and the first by any botanist not based on a visiting ship, it has served as a major source for nomenclature for nearly two centuries.

Anyone with any familiarity with New Zealand and its botany, on reading the text of the Precursor as a whole rather than by picking out individual species, would become well aware of the large number of typographic errors which mar the work. These errors have generally been corrected, often silently, by past authors.

In recent times, there has been a tendency to insist on the orthography of taxonomic names as they were originally printed, even when the author himself or herself has later explicitly labelled them as an error and corrected them, and the correction has long been accepted. This tendency has been described as unfortunate, tedious, and even perverse (Brummitt 2011, pp. 1205–1206).

Part of the problem is that the International Code of Nomenclature for algae, fungi and plants (hereafter, ICN; McNeill et al. 2012), while allowing the correction (even silently) of errors, does not completely specify what constitutes a typographic or orthographic error which it is permissible to correct, apart from specific cases in Latin and Greek. ICN Art. 60 includes examples of a correction of an orthographic error being allowed (Ex. 5) and not allowed (Ex. 8).

An autograph manuscript of the Precursor still exists, and by reference to this it is possible to confirm that most of the errors that appear were made during the transfer from the manuscript to the printer’s type. And in any case, most of these errors are fairly obvious and trivial, and without nomenclatural consequences.

There is one typographic error which stands out immediately: in the title of the work, Cunningham’s manuscript has the correct Latin spelling Praecursor, rather than the English word Precursor which heads all parts of the printed version. The spelling has occasionally been silently corrected (e.g. Cockayne 1910, p. 17), but that will not be followed here.

However, there are two errors which have been taken up by recent authors and have made their way into significant taxonomic publications. There are also two other nomenclatural ones which have remained more or less unnoticed, and in any case which have little or no potential significance.

The most important error is the spelling of the fern genus Loxsoma, printed in the Precursor as Loxoma (Cunningham 1837, pp. 366–367, pls 31–32). This was labelled an error and corrected by Hooker (1838), and the spelling Loxsoma has been accepted since that time, until the influential taxonomic work of Tryon & Tryon (1982) deprecated that spelling, without citing any specific justification. Since then, a number of other significant works have followed the Tryons. For example, the spelling of Loxsoma and its family name Loxsomataceae in the otherwise excellent work of Smith et al. (2006) as ‘Loxoma’ and ‘Loxomataceae’ is in direct contradiction to their claim to have used well-established names, as the authorities they cite (e.g. Hoogland & Reveal 2005) do not accept those versions.

As a result of this confusion, Alfarhan et al. (2009) proposed to make Loxsoma R.Br. ex A.Cunn. a nomen conservandum with that spelling. This proposal was accepted by the IAPT Nomenclature Committee for Vascular Plants (Brummitt 2011, p. 1205).

The situation with Loxsoma is actually not simple, and is far from what was stated by Alfarhan et al. (2009). This case will be examined in detail in this article, with the result that the author of Loxsoma and its species is not Cunningham, as has been accepted in the past, but Sir William Jackson Hooker. This is actually stated in the printed text of the Precursor, but has been overlooked by all previous workers.

The other three errors will also be examined; by reference to the autograph manuscript these are seen to be cases of simple typographic errors. It should be noted that no errata relating to the Precursor were published in either of the journals in which it appeared.

A further valuable reference is the recently published correspondence of Allan Cunningham (Orchard & Orchard 2015). This gives a great deal of background information on the substantive matters discussed in this article.

Background

Allan Cunningham had collected specimens in New Zealand during the period August–December 1826 while an agent for Kew in the Australian colonies. He returned to England in 1831. His brother Richard Cunningham had also collected in New Zealand for several months in 1833–1834 while Colonial Botanist in New South Wales. A few of the plants collected by them were published in some of Hooker’s journals with descriptions by Allan Cunningham (e.g. in Hooker 1836), but the majority remained unpublished.

Richard Cunningham was killed during an expedition in 1835 to the Bogan area of New South Wales. His position was offered to Allan, who left England at the end of October 1836 (Hooker 1837). After resigning as Colonial Botanist in 1838, Cunningham then travelled again to New Zealand to extend his collections at his own expense, and returned to Sydney in poor health. He died there on 27 June 1839 (Heward 1842).

Genesis of the Precursor

While Allan Cunningham was living near Kew during 1831–1836, Hooker during that period was resident in Glasgow. On a visit to London, Hooker saw and was apparently impressed by a manuscript list of the known flora of New Zealand that Allan Cunningham had prepared for use by his brother Richard. This was in an abbreviated format unsuitable for publication (Cunningham, letters to Hooker in Orchard & Orchard 2015, p. 347, 349). After being appointed Colonial Botanist in about June 1836, Cunningham immediately set about preparing for an imminent prolonged or even permanent absence from England. One of many tasks was to revise the manuscript, including examining Banks’s herbarium in London, ‘a considerable operation & at a period when every Moment ought perhaps to have [been] otherwise Employed’, and write an introductory essay ‘in great haste—for I have had to steal an hour at night whenever I could’ (letter to Hooker in Orchard & Orchard 2015, p. 349).

On 23 September 1836, Cunningham at last was able to post the finished manuscript to Hooker. His covering letter (in Orchard & Orchard 2015, pp. 349–350) stated:

[T]he names for your notice you will be pleased to attend to … Dr. Greville will assist you and regarding Loxoma and Bauer’s beautiful drawing I must particularly [ask] that you will do justice to so remarkable a plant … I enclose … a Marchantia looking plant which you will be able to make out better than I can, and fill up the blank in Mss p. 29.

The significance of these remarks will be made apparent below.

Although Hooker was in Glasgow, his printers were in London. This had led to problems in the past. Cunningham had previously complained several times about the number of errors appearing in his papers in Hooker’s journals. He expressed doubt that Hooker was checking the proofs, surmising that ‘after the mss has passed your hands … you don’t see any more of it, no proof being sent to you’ (letter to Hooker in Orchard & Orchard 2015, p. 316). Eventually he received permission to go and correct the proofs directly with the printers (letter thanking Hooker in Orchard & Orchard 2015, p. 325). This was obviously impossible with the Precursor, publication of which began after he left England, and continued after he died.

The manuscript

An autograph manuscript of the Precursor is still held by the Royal Botanic Gardens Archives, Kew, and microfilmed copies made by the Australian Joint Copying Project are available in some overseas libraries. This study utilised a microfilm copy from the National Library of New Zealand (Cunningham 1836).

The manuscript cannot be described as a fair copy fit for the printers. Many of Cunningham’s papers show considerable numbers of corrections, but the manuscript of the Precursor is the worst seen. Except in the introductory historical essay, it has numerous overwritten corrections, and extensive crossing out of whole species descriptions for which the correct text is on tipped-in slips of paper. Some species were inserted late, leading to extensive renumbering. There are also quite a few small sketches of plants or parts of plants, none of publication standard, and all crossed out. Furthermore, Cunningham’s handwriting is rather compressed and angular, making it often difficult to read.

This would be a compositor’s nightmare. But this manuscript is very unlikely to have been the copy used by the printers. It does not contain material added by Hooker, which is discussed below. It could, however, be the copy sent to Hooker, for a number of reasons.

Firstly, Cunningham was pressed for time, as is obvious from comments in his letters from this period, some of which have been quoted above. He probably did not have the time to transcribe an error-free copy. If he had, he would likely have attached the introductory essay to that copy, as that portion of the existing manuscript is quite legible and has few corrections. Secondly, the manuscript contains an index (which included Māori and English common names) with page numbers, which is referred to in his covering letter. It must be asked whether he would have spent the time to construct an index for a rough draft. Finally, it contains a memo to Hooker with exact details of which parts of the manuscript required additions. This was the ‘names for your notice’ mentioned in his covering letter, as quoted above.

Possibly the copy actually sent to the printers was a transcription of the present existing one with Hooker’s additions. The transcriber may have been Hooker, or—more likely given the many errors introduced into the Latin portions—an amanuensis in his employ. Given such source material, it is no wonder that there should be numerous typographic errors.

The manuscript of the Precursor bears many later marginal annotations by Joseph Dalton Hooker, including revisions of the nomenclature which appeared later in his Flora Novae-Zelandiae (Hooker 1853), and new Māori names provided to him by ‘Col[enso]’. Nevertheless, Hooker held a poor opinion of the Precursor, which for some unknown reason he habitually cited as ‘Prodromus’, a term not found in any of Cunningham’s published or unpublished work. According to Hooker, ‘the work is so unsatisfactory and incomplete that were it not for the invaluable herbarium of both Cunninghams … I should have found it impossible to have quoted the “Prodromus” with any degree of confidence’ (Hooker 1853, p. iv).

The memo to Hooker

At the end of the existing manuscript is a one-page memo beginning ‘Sir Wm. J. Hooker is requested to attend to the following blanks and corrections in this Mss.’, signed with Cunningham’s initials.

Four of the items in this memo requested Hooker to obtain published descriptions from works to which Cunningham did not have access at that time, for example those of some French voyages (Bory de St. Vincent 1828; Bélanger & Bory de St. Vincent 1834). Of these, Hooker was able to supply only the generic description of Earina from Lindley (1834), which was based on Cunningham’s specimens.

Three other items were for species based on manuscript names by Robert Greville. For these, Hooker was requested to ask Greville to supply descriptions. He did so for two seaweeds, and Greville’s name appears after the specific descriptions of Rhodomenia [Rhodymenia] lusoria and Thamnophora cunninghamii. By ICN Art. 46 (cf. Ex. 22), the authorship of these names is attributed to Greville, not Cunningham, and this is well known in the phycological community (e.g. in the databases at http://www.algaebase.org/search/species/).

One other item was to determine the species of Marchantia represented by Cunningham’s specimen and insert it into the text. Hooker apparently determined this as M. polymorpha L.

The most important item, however, was this:

Loxoma. Prof. Hooker is particularly requested to draw up elaborately a generic char[acte]r. of this rare plant, specimens in his own herbarium probably as also the beautiful drawing from the pencil of the inimitable Bauer, sent in this Mss. affording him ample means.

The consequences of this will be more closely examined in the next section.

As shown by his correspondence (Orchard & Orchard 2015), Cunningham was frequently in contact, either by letter or in person, with Greville, Lindley and Brown. His asking Hooker to obtain information from the first two, and to provide a diagnosis for Brown’s new genus, is further evidence of how pressed for time he was.

The case of Loxsoma

Loxsoma cunninghamii was originally discovered by Allan Cunningham in 1826. Puzzled as to its classification, he originally thought it a species of Trichomanes, and under that name forwarded specimens to Hooker.

The evolution of the name is traceable in a manuscript headed ‘List of specimens collected in New Zealand’ (Cunningham 1827). Here, it appears first as Trichomanes sp., then as T. conopteroides. These are crossed out and Davallia dealbata is written in. Later, the name Loxsoma Cunninghamii has been added, probably by someone else. This list appears to be the source of at least the other manuscript names mentioned in the Precursor.

Cunningham’s description in this manuscript list is:

A fern of robust growth, in habit like Conopteris but with marginal sori, & capsules inserted, a cylindl. receptacle with an urceoliform involucrum. This remarkable fern I detected in small wood at Waterfall near Kiddee Kiddee. Fronds laciniated white beneath.

In the manuscript of the Precursor there is no species description, only a generic one. This is headed ‘Loxoma. R. Br. Mss. ined.’, with a date, 1835, crossed out. The description is:

Involucrum subsuperficiarium monophyllum urceolatum ex [illegible] exterior texturâ frondis. Capsula pedicel[illegible] verticaliter dehiscentes receptaculo communi inserto. Annulus latus incompletus obligem. A genus lately founded by Mr. Brown on the structure of the capsular ring which is remarkably oblique. Its affinity is with Dicksonia and Trichomanes. Filix glabris bipedalis et infra. Stipes compresso-teres, fuscus. Frons pedalis deltoidea bipinnata. Pinnula pinnatifida lanceolata attenuata, supra viridi, subtus dealbata.

The abbreviation ‘Tab.’ next to the species name ‘L. Cunninghamii Br. mss. ined.’ indicates that Cunningham definitely intended Bauer’s figures to be printed. Attached to this is a tipped-in slip of paper which reiterates the request in the memo:

Loxoma. Sir Wm. Hooker is requested to draw up [illegible]elaborately the character of this new & beautiful genus of Mr. Brown’s, from the splendid & highly magnified details furnished Allan Cunningham by his ex[c]ellent friend Mr. Bauer, & here sent to Profr. Hooker, [illegible] specimens of which A.C. presumes he has in his herbarium.

A comparison of the above descriptions with those printed in the Precursor shows that Hooker had carried out Cunningham’s request, and that the whole of the printed protologues for both Loxsoma and its species L. cunninghamii are entirely the work of Hooker. Moreover, this is clearly shown by Hooker’s name printed in parentheses after this section of text, following comments to that effect (‘I have [drawn] up the above generic character’). The provision of the author’s name in parentheses after a section of text was a common method of unequivocal attribution in works of that period, including those of Hooker (see, for example, ICN Art. 46 Ex. 22). As noted above, this explicit attribution, which satisfies the requirements of ICN Arts 46.6 and 46.8, has been completely overlooked by later workers. As the name originated from Robert Brown, the correct spelling would be whatever Brown suggested. This therefore seems to be a case of an orthographic error in the transmission from Brown via Cunningham to Hooker.

No relevant material appears to be present in the surviving papers or correspondence of Brown. Other evidence seems to show that, fairly soon after publication, the correct spelling was made known to both Cunningham and Hooker. This correction was evidently received by Cunningham after his return from New Zealand. In a letter to John Smith written in New Zealand on 20 May 1838 (in Orchard & Orchard 2015, p. 460) he still uses the spelling ‘Loxoma’. After his return, he mentions being informed of the correct spelling by Brown, in letters to Hooker dated 28 November 1838 and to William Colenso dated 24 October 1838 (in Orchard & Orchard 2015 p. 442, 462).

A later Cunningham manuscript titled ‘Florae Novae Zelandiae’ exists (Cunningham 1838c). This is a list of species, based on the manuscript of the Precursor, as shown by the seaweed manuscript name Rhodomenia praelonga, which Greville changed to R. lusoria in the printed version, as noted above. The text also contains notes added during his visit to New Zealand in 1838. In this manuscript, Cunningham originally wrote ‘Loxoma’ but added an ‘s’ as an insertion. There is a similar insertion in the names on the labels of a specimen (now at Kew: K000913881) in that portion of his herbarium which Cunningham took with him when he moved back to Sydney.

Finally, there is the correction of Hooker (1838), who admitted ‘by an error it is printed Loxoma’. The date of publication for this was about 1 May 1838 (Stafleu & Cowan 1979, p. 296), therefore certainly prior to Brown’s communication reaching Cunningham. This is a case of a publishing author correcting his own error, and doing so in a reasonably timely manner. The proposal of Alfarhan et al. (2009) may have been rejected as superfluous had this been known by the Committee, because they did reject just such a similar proposal (to conserve Mitrastemon Makino with that spelling; Brummitt 2011, p. 1206). However, there is no evidence that the Committee was aware of this.

The correct full citation for the genus is therefore Loxsoma R.Br. ex Hook. in A. Cunn. Companion Bot. Mag. 2: 366. 1 July 1837 (‘Loxoma’), while the species is L. cunninghamii R.Br. ex Hook.

The erroneous attribution to Cunningham undoubtedly stems from the importance formerly given to the person who first recognised and named a new taxon, in this case Brown, even if they did not publish it. It was not until the Cambridge Congress of 1930 that the author who first published the diagnosis or description was considered to have more importance (Rendle 1934, art. 48). By then, it seems, workers had ceased to consult the original work and relied on later (but pre-1930) synopses, floras and indexes.

Even Hooker himself, in a work published in his old age, stated the ‘genus … was first discovered and described by Mr. Allan Cunningham’ (Hooker 1862, t. 31). This perhaps indicates a failure of memory; in the same work he states ‘Presl seems to have taken no notice of it’ and instead attributes the family ‘Loxsomaceae’ to the recently published paper by van den Bosch (1861). Presl (1847, pp. 31–32) had in fact validly published the family many years before, after discussing the need to create ‘einer neuen Abtheilung oder Tribus’ to hold the genus rather than accept Hooker’s placement of it among the Davalliae (Presl 1843, p. 6).

Etymology

Because of the lack of surviving material by Brown, there is some question about the etymology of the name Loxsoma. It is generally agreed that it is derived from two Greek words, the first of which is λοξος, oblique or inclined. The interpretation accepted today (Brummitt 2011, p. 1206) is that the second word is σῶμα, body, and that it refers to the sporangia. The first mention of this appears to be by Hooker (1855, p. 18). Cunningham, however, seemed to think the second word was λῶμα, belt, referring to the annulus. This is the derivation he wrote on the label of the herbarium specimen mentioned above, and which he also communicated (the English word only) to William Colenso (letter cited above, in Orchard & Orchard 2015, p. 462). As the ‘s’ in the name Loxsoma written next to the Greek on the herbarium label is evidently a later insertion, this derivation quite possibly pre-dates Cunningham’s knowledge of the correct spelling.

Which derivation is correct is a matter for conjecture. On one hand, it could be assumed that Cunningham received his version from Brown. On the other, it could also be assumed that Joseph Hooker may have received his either from his father or directly from Brown. It is regrettable that there is no current evidence either way. It can only be remarked that both Cunningham’s and Hooker’s descriptions only mention the oblique annulus, not inclined sporangia. In any case, it is evident that Loxsoma is a portmanteau word, not a simple compound.

It should be noted that both possible endings are from neuter nouns ending in –ma and with genitive singulars ending in –matos: under the current ICN, Art. 18.1, the spelling of the family name as Loxsomataceae, which Pichi-Sermolli (1993 and references therein) spent so much effort to justify, is now automatic.

Typification

Loxsoma cunninghamii was lectotypified by Allan (1961, p. 22) (who wrongly credited Endlicher with being the author of the correction to Loxsoma). Brownsey & Perrie (2015) more precisely identified the specimen chosen as being in the John Smith Herbarium, Natural History Museum, London, and as having a pencil drawing of the sori on the sheet. This corresponds to a specimen with registration number BM000905149. Pat Brownsey (Museum of New Zealand Te Papa Tongarewa, pers. comm., 2016) now considers that this choice is not valid, as the specimen would not be one of those in Hooker’s possession when he wrote the protologue. Presumably, Hooker’s herbarium came to Kew after he moved down from Glasgow. However, the limited number of Kew specimens for which information could be obtained during this study were either in Cunningham’s possession when Hooker wrote the diagnosis, or were collected by Cunningham during his 1838 visit to New Zealand. Consequently, it has not been possible during this study to determine a suitable replacement lectotype.

Other errors

One further error has made an appearance in the taxonomic literature. This was for the genus Griselinia G. Forst., which was printed as Griseliniae (Cunningham 1839c, p. 261). This caused Reveal & Doweld (1999) to assume that Cunningham was proposing to create a family Griseliniae based on some Forsterian name. It has already been pointed out that not only is this assumption invalid under ICN Art. 37.2, but it is in any case untenable given the surrounding context (Earp 2013). The autograph manuscript of the Precursor confirms that Cunningham wrote simply Griselinia.

Two other simple typographic errors affect species names but are too trivial to have been taken up by other authors. The first was printed as Lagenophora fosteri (Cunningham 1838b, p. 125), although the manuscript shows that Cunningham wrote forsteri. This was intended as a replacement name for Calendula pumila G. Forst. Cunningham was unaware that L. forsteri had already been proposed as a replacement name for Forster’s species, published in the same month he left England by De Candolle (1836). This is shown by the fact that in a later manuscript (Cunningham 1838c) he attributes the name to himself. Indeed, it is possible that specimens so labelled had been among those he had sent to De Candolle, knowing the latter was working on a volume on the Asteraceae (letters to De Candolle, in Orchard & Orchard 2015, pp. 399–402). The first correction was probably by Raoul (1846) who still attributed it to Cunningham, the author then being corrected by Hooker (1853, p. 125). Regardless of the author, the name L. forsteri is illegitimate under ICN Art. 11.4, as it does not carry over Forster’s specific epithet. The species is now known as L. pumila (G. Forst.) Cheeseman.

The second was a new species of Pittosporum, with specific name printed as engeinoides (Cunningham 1839c, p. 106). This has, ever since, been known as P. eugenioides, a name which came close to losing priority (Cooper 1956, p. 181) because of Hooker’s decision to spread the publication over so many months. The manuscript of the Precursor shows that the latter spelling is correct, but it also shows that it is entirely understandable that Cunningham’s handwriting could be misread in this case. Apart from his ‘u’s looking like ‘n’s (the source of many misspellings in the printed version), a spot of ink appears above the ‘n’, looking like the dot on an ‘i’, whereas the dot above the following ‘i’ is fainter. The correction, first published by Raoul (1846), is fairly obvious, given that the printed description in the Precursor contains the words ‘aspectu fere facie Eugeniae’.

An interesting case, which does not require any correction, is Persoonia toru, the specific name of which appeared in the printed Precursor as Tora (Cunningham 1838a, p. 378). William Colenso complained about this in a letter to Hooker (in St. George 2009, pp. 196–197):

Allow me, also, to request that you will be pleased to turn to Cunn.’s Ms., for the specific name of his N.Z. Persoonia which cannot (must not) be ‘Tora’ [priapic] (a most obscene word); Toru is the Native name of the tree, and Cunningham, who had all the names either from, or corrected by, the Missionaries, – must have written it Toru. If you find it to be as I suppose, you can easily alter it; and if not, do try to change its nom. sp.

Fortunately, this species had been validly published prior to the Precursor with the correct spelling (Cunningham in Hooker 1836), presumably because Cunningham had then been on hand to correct the proofs.

Apart from these, there are numerous other typographic errors which are quite obvious. Some examples are ‘Acoena’ (Acaena), ‘Aledryon’ (Alectryon), and ‘Vittaclinia’ (Vittadinia). Even the descriptive Latin text is not immune. Cunningham’s new genus Ixerba is said nonsensically to be in (‘intus’) two different families (Cunningham 1839b, p. 249), rather than between them (‘inter’) as in the manuscript. This suggests that any use of Cunningham’s descriptions should be verified with the manuscript version of his text.

Finally, there is a lacuna in the printed version, whereby the generic heading for Bryum is followed by the description of an ‘H. pinnata’, and the species numbering jumps from 97 to 100 (Cunningham 1837, p. 335). The manuscript supplies what is missing: entries for Bryum bartramioides and B. dichotomum, and a generic heading for Hookeria, with the following species correctly spelt as H. pennata.

Conclusions

The printed version of Cunningham’s Florae Novae Zelandiae Precursor contains numerous typographic errors. Most of these are in general text and would only be noticeable to a person familiar with New Zealand, such as the substitution of ‘n’ for ‘u’ in Māori words, particularly frequent as Cunningham used ‘u’ in his manuscript where often the modern spelling would have ‘w’. The few that affect taxonomic names are trivially correctable, but errors in the general text may not always be so noticeable.

However, the spelling of the generic name of Loxsoma as Loxoma is an orthographic error, not a typographic one. It was corrected very soon after publication by the actual publishing author of the name, Sir William Hooker, and this correction should perhaps have been allowed to stand, instead of requiring to be listed in the ICN nomina conservanda. Hooker’s status as author of the name should have been clear enough from the unequivocal attribution in the published protologue. The manuscript version of the Precursor removes any doubt and supplies an explanation, although under ICN Art. 46.8 it cannot be taken into account as formal evidence.

When a work is as obviously flawed as the printed version of the Precursor, there is a need for caution when either simply citing it, or when using it as a basis for more advanced nomenclatural or taxonomic expositions. This warning was originally given by Hooker (1853), but seems to have gone unheeded since.

Acknowledgements

I wish to thank the referees for their comments, which have materially improved the manuscript. Thanks also to Alison Paul (Natural History Museum, London) and Pat Brownsey (Museum of New Zealand Te Papa Tongarewa, Wellington) for supplying material on the current lectotype and other specimens.

Associate Editor: Dr Rob Smissen.

Disclosure statement

No potential conflict of interest was reported by the author.