Malaxideae (Orchidaceae) in Madagascar, the Mascarenes, Seychelles and Comoro Islands

Summary. The tribe Malaxideae (Orchidaceae, subfamily Epidendroideae) (sensu Pridgeon et al. 2005: 453) in Madagascar and adjacent archipelagos is revised. In this region it comprises the four genera: Liparis, Malaxis, Oberonia and Stichorkis. All of the species are described and their typification, history, identification, distribution and habitat are discussed. Conservation assessments and distribution maps are included. A checklist of the genera and species and a key to their identification are provided. Six new species: Liparis bemarahensis, L. bosseri, L. chantaliae, L. laurentii, L. magnifica and L. superclareae are described.


Introduction
This is the first review of the tribe Malaxideae (Orchidaceae, subfamily Epidendroideae) from the Madagascar region since 1936 when Henri Perrier de la Bâthie (henceforth Perrier) published his Les Liparidinées de Madagascar, which was later summarised in his treatment of the Orchidaceae in the Flore de Madagascar . His work covered the same genera but only from Madagascar and the Comoros, the present revision also adds the adjoining archipelagos of the Mascarenes (Réunion, Mauritius, Rodriguez) and the Seychelles. Historically, accounts of the orchids of the area under consideration have been fragmented with revisions being undertaken separately for Madagascar, the Comoros, the Mascarenes and the Seychelles. This sometimes resulted in confusion with the same species being known under different names within its range: thus Liparis scaposa Frapp. ex Cordem. (Frappier in Cordemoy 1895: 183) described from Réunion is identical to Liparis microcharis Schltr. (Schlechter 1924: 144) from Madagascar, as shown below. This account is also the first since Ridley's monographs (1887Ridley's monographs ( & 1888 where the species from the region were compared with those from mainland Africa and elsewhere. Increased collecting in some parts of Madagascar and improved communication have meant that today much more material is available for research in herbaria around the world. This is in contrast to the time of Perrier's revision when many of the species were only known from their type collection.
The large genus Liparis has caused problems for earlier authors and disproportionate numbers of plants have been either misidentified or left unidentified in herbaria. This is not surprising as many of the species have few good and consistent characteristics. Plant habit, lip morphology, shape of the column and shape of the anther can all vary to some extent within a species but rarely are any of their individual characteristics unique to a single species. Therefore, one must often rely on a combination of features that are not always obvious. Frustratingly, in herbarium specimens the lip of the flower is often damaged, distorted or has lost some of its defining characters, the anther cap is also frequently missing.
The typification of 18 th and early 19 th century species can be challenging because specimens were commonly moved or incorporated into other herbaria. Frappier (1880) and Frappier in Cordemoy (1895), working on the orchid flora of Réunion, was vague as to which collection or collections were the basis for his new species. Schlechter and Perrier, in their descriptions, frequently cited several collections, each often comprising several herbarium sheets. Therefore, where necessary, a careful selection has been made in this paper to assign a lectotype or new material was assigned as a neotype to best reflect their descriptions. Some of Perrier's earlier collections followed a different numbering system from that used by Schlechter for the same specimens. In addition it was once thought that the material sent by Perrier to Schlechter in Germany was not returned and was lost during World War II and therefore most of the Paris specimens would have been isotypes. However, the loan and return dates of herbarium sheets are indicated very clearly on various herbarium labels and it is therefore certain that the holotypes and other material was returned to Paris (e.g. Perrier 14358, 16485, 15336 etc. in P). For clarity and following the International Code of Nomenclature (Turland et al. 2018) and also taking into account McNeill (2014McNeill ( : 1112 for species described prior to 1958, holotypes were recognised when one particular specimen was indicated by the author in the protologue as the type and/or used by the author when no type was indicated (article 9.1); for example Perrier and Schlechter generally annotated the Paris (P) herbarium specimens as 'type', thereby clearly indicating the particular original material used by them (article 9.4). In the case of more than one sheet of original material with the same collecting number being available, the identity was checked and the most representative sheet designated as lectotype and the others as isolectotypes or isotypes (article 9.5). If additional collections of original material were mentioned in the protologue and found to be correctly identified and used by the author they were considered syntypes (article 9.6) or lectoparatypes following the instructions for this journal. On a few occasions neotypes had to be selected when original material is missing or destroyed (article 9.8).

Methods
This revision is based on detailed work during the last 25 years on the orchid flora of Madagascar and neighbouring islands. J. & C. Hermans (in Du Puy et al. 1999) published a comprehensive bibliography which was revised  and has been kept up-to-date since then. At the same time, photographs, drawings and data were assembled from herbaria (including BM, BR, BRLU, CGE, DUB, G, HEID, K, M, MO, P, SZU, TAN, TEF, W, WU, M, SZL, ZSS). The authors visited many type and other localities during extensive fieldwork and visits to Madagascar and Réunion.
A new checklist was compiled, based upon examination of herbarium and photographic material in all the major collections. Species were typified or retypified as necessary, a provisional key to the species was also prepared. All the unidentified herbarium specimens and photographic records were then attempted to be identified and named, some being recognised as novelties. A few proved to be impossible to identify because of their poor condition or missing parts. New descriptions, including information on distribution, habitat, elevation etc. were prepared for all the species. The keys were tested on further herbarium and spirit material made available by Patrice Bernet and others from Réunion and on the extensive collections provided by Simon Verlynde and Tariq Stévart from the Ambatovy orchid conservation project in Madagascar. Meanwhile, new botanical drawings were commissioned by the Royal Botanic Gardens Kew to illustrate novelties and other significant species not illustrated before. Simon Verlynde with input from Brigitte Ramandimbisoa provided the detailed preliminary Conservation assessments. Jean-Michel Hervouet extracted the locality information and constructed distribution maps for each species.
As mentioned above, it was often necessary to rely on a combination of characters to identify many of the species in this tribe: plant habit, shape and features of the lip, shape of the column wings and anther cap are only reliable characteristics for species when used in combination. The key features for each species are emphasised in the descriptions and especially in comparison with similar species. Many Liparis are very variable in respect of the number and lay-out of the floral veins and these characteristics were therefore not included unless significant or unusual. The shape and size of the lip and the calli on the surface can also be variable. These variations are both within individual plants and between different plants; for example, in Liparis ochracea Ridl. (Ridley 1885: 461) the upper flowers are sometimes incomplete, irregular and misshapen. A species can vary locally but the differences are not enough to recognise them as distinct taxa. In some cases it is possible that they are developing into new species but it is difficult to be certain how these sporadic, isolated variations have evolved.
Only the most relevant illustrations and literature are mentioned in this revision. References to additional literature and illustrations for all the species can be found in the annotated bibliography in . Photographs and / or drawings or references are provided for all species accepted in this revision. Because of the general confusion and often misidentification of specimens of Malaxideae in herbaria, all specimens that we have examined are listed.
Type specimens are only indicated for those names from the region under review and not for synonyms from elsewhere. Most of the Cordemoy herbarium material has been on long-term loan to P and later to REU but belongs to the Herbarium at the St. Charles University, Marseille (MARS), it has therefore been indicated as such (also see Bosser 2011: 118).
Since there is a lack of detailed genetic research, no distinct sections in the genus Liparis are accepted in this paper. It is evident that some species cannot be assigned easily to the sections designated by previous authors. They overlap with the classical sections Mollifoliae and Coriifoliae used for Liparis in Africa and Asia. For identification purposes (and in the key) three broad, overlapping groups are recognised for Liparis: 1. Species with pseudobulbs often caespitose, the older ones retaining their leaves. Leaf blade more or less flat, rigid or leathery. 2. Species with few (one to two) pseudobulbs, the older ones disintegrating. Only the newest growth bears leaves; leaves plicate and thin. 3. Species with very small, often clustered pseudobulbs. Plants very small to small. With small rounded pseudobulbs and ligulate-lanceolate leaves.
Unlike some descriptions in the literature, overall plant size given does not include the inflorescence. Inflorescences are often disproportionate to the plant and far more variable than the vegetative parts of the plant, it was therefore considered a better characteristic.
Unless otherwise stated, all cited specimens have been seen by the first author. Herbarium barcodes (if existing) were only indicated if no collection numbers were available or if there was a possibility of confusion.

Conservation status
Using the IUCN Red List Categories and Criteria (IUCN 2012), we produced preliminary risk of extinction assessments for each species. In order to gather the most comprehensive georeferenced distribution dataset, historical collections were georeferenced post facto when locality information was sufficiently precise and if not, these specimens were not taken into account in the preliminary assessments. We imported georeferenced specimen data into the dedicated R package, ConR (Dauby et al. 2017) to calculate the Extent of Occurrence (EOO) and Area of Occurrence (AOO). Minimal AOO was calculated using a cell area set to 4 km 2 (cell size = 2 × 2 km) as recommended by IUCN (2017). The number of locations (sensu IUCN) were estimated with regard to the size of the main threat, in which a single location may encompass more than one adjacent subpopulation. For these preliminary assessments, only the criterion B, based on three parameters, EOO, AOO and number of locations thresholds, was used. These preliminary assessments should not be considered as IUCN Red List assessments, but as reliable indication of the threat level weighing on these species. (S. V.).
The assessments showed a considerable number (15%) of Malaxideae species in the area are Critically Endangered (CR), the vast majority (47%) are Endangered (EN), 18% are Vulnerable (VU) and only 18% are of Least Concern (LC). The main threats are habitat loss through slash-and-burn farming, logging, charcoal production and mining.
A number of species are relatively common (i.e. known from ten or more different collections) but others are quite rare and localised (fewer than ten collections): including Liparis ambohimangana Hermans (Hermans et al. 2007: 214), L. bathiei, L. bemarahensis, L. bosseri, L. cladophylax, L. danguyana, L. densa, L. gracilipes, L. laurentii, L. lutea, L. magnifica, L. parva, L. vulturiceps, L. zaratananae Schltr. (Schlechter 1924: 151) and all Malaxis except M. seychellarum. A few are known from the type specimen only: Liparis trulliformis Schltr. (Schlechter 1924: 149) and Malaxis madagascariensis, they may well be reassigned to other species when more thoroughly studied using better material. The newly described Liparis superclareae is only known from the type collection which comes from a remote area. The overall distribution of the Tribe is shown in Map 4.

Distribution mapping
Georeferenced maps with WGS 84 projection have been used as a background for the distribution maps. They have been freely downloaded from the following websites: www.fao.org/geonetwork/srv/en (land cover of Madagascar, dated 2009), https://www.data.gouv.fr (Mayotte and Réunion islands), https://gadm.org (country outlines of Comoros, Mauritius and Seychelles). Similar data can be found at www.diva-gis.org. The locality coordinates come from the authors' own observations, made with Garmin etrex 10 GPS using the WGS 84 projection, or from the Tropicos website gazetteer to Malagasy botanical collecting localities: www.tropicos.org. All the positions of specimens have been double-checked with Google Earth: https:// earth. google.com. A relational database with two tables: localities and observed species, has been built with the database management system 4D version 11 (www.4d.com), storing a total of about 500 observations or collected specimens. A computer program was then written in 4D programming language to automatically produce one CSV file per species (ASCII files with commaseparated values), every file containing a list of coordinates for a given species. Eventually the open source Geographical Information System QGIS version 2. 18.10 (www.qgis.org) was used to compose the final pictures in jpeg format, combining the countries' outlines, Madagascar land cover and the location spots for every species. Distribution is only shown for the area under review in this paper.
A simple underlying vegetation map was used as a background for each species, a map of the whole area covered and the distribution of all Malaxideae, as well as maps showing the different provinces and regions of Madagascar are included for reference (Maps 1,2,3 and 4).

Phylogenetics & Cytology
The Malaxideae tribe has always been difficult to place within the Epidendroideae subfamily because the included taxa lack typical appendages on their pollinia (e.g. they are 'naked', although they appear to have small viscidia in most cases). Column morphology, particularly the structure of pollinia and associated appendages, has always been given pre-eminence in traditional orchid classification, so ideas about relationships within this 'nearly naked' tribe have been limited. In all other aspects Malaxideae are more or less typical epidendroid orchids without any clear evidence for any relationship to any other tribe. In the morphological analysis of Freudenstein & Rasmussen (1999), two species of Liparis were among a large number of Epidendroideae on a polytomy, whereas in the Cameron et al. (1999) analysis of rbcL data, Liparis and Malaxis were sister to Dendrobieae but without bootstrap support greater than 50%. This placement is noteworthy because both groups have more or less naked pollinia, but Dressler (1993) hypothesised that whereas Dendrobieae appear to be secondarily naked, Malaxideae may be primitively so. As in the case for most clades of Epidendroideae, more data are required before a clearer picture of the relationships of Malaxideae is achieved.
A comprehensive account of Malaxideae has yet to be published although an up-to-date assessment of its phylogeny and classification was provided by Chase & Cribb in Genera Orchidacearum (Pridgeon et al. 2005). Cameron (2005) examined a selection of species and produced a cladogram which shows that both Liparis s.l. and Malaxis s.l. are polyphyletic. Very few plants from Africa and the Madagascar region were included in this study and this was recognised by the author: 'Eventually, the tribe will need to be reclassified, but it may be premature to carve Malaxideae into more genera at the present time because fewer than 10% of the tribe's species have been sampled, and the circumscription of new genera would be somewhat speculative. Substantially more taxa (especially from Africa and the Neotropics) should be sequenced before a thorough revision of this troubled tribe's taxonomy is attempted. Nevertheless, if the current well-supported pattern of species relationships continues to be recovered in larger analysis, it may be that Malaxideae can be reclassified at the genus level based first on habit and vegetative features, with floral characters playing a secondary role. If this is ultimately found to be the case, it will be a remarkable exception within Orchidaceae and angiosperm taxonomy, in general, where the flower above all other organs reigns supreme. ' Szlachetko (1995) recognised 16 genera in subtribe Malaxidinae, seven newly described by him. Szlachetko & Margonska (2002) accepted 15 genera in the subtribe, two of them new, in their subtribe Oberoniinae. They provided detailed drawings of the gymnostemium structure of representatives of these genera. Senghas (2002genera. Senghas ( : 2710 added new genera to his treatment in the third edition of Schlechter's Die Orchideen and provided a tabular comparison of some of the segregate genera of Malaxis recognised by Szlachetko & Margonska. Margonska & Szlachetko (2004) established another segregate genus, Disticholiparis, for those species previously referred to Liparis sect. Distichae.
We remain to be convinced that all the new segregate genera of Malaxis and Liparis warrant recognition at more than infra-generic rank, as already recommended by Ridley (1888), Smith (1909Smith ( , 1913, Schlechter (1911aSchlechter ( , 1915b and Seidenfaden (1978). Salazar (personal communication) and Cameron (2005) have suggested on the basis of DNA sequences that the New World taxa of Malaxis and Liparis may be more closely related to each other than to the Old World species in those respective genera. Cameron (2005) identified four major clades in the tribe that are characterised by their habit and leaf type: recognise 23 genera within two subtribes in Malaxideae. (Adapted from Cribb & Chase in Pridgeon et al. 2005: 453) According to Felix & Guerra (2010: 245) the Malaxideae tribe has a great diversity in chromosome numbers, with n = 21 predominating in Liparis and Malaxis and n = 15 in Oberonia. The most probable basic numbers for the tribe are x = 15 and x = 21, n = 21 being the most frequent haploid number and n = 15 the second most frequent and the only one that occurs in all three genera. The occurrence of n = 14 in some species of Liparis and Malaxis, and the high frequency of n = 21 in these two genera, suggest that these numbers may be related to the polyploid series x = 7, 14, 21 that is quite common in orchids. Oberonia has n = 15 or 30. The occurrence of n = 15 in Malaxideae further supports the hypothesis that the karyological evolution of the family was principally through dysploidy of approximately one chromosome in the three principal ploidy levels, n = 7, 14, 21 ± 1.
Recent observations by one of the authors (J. H.) in the Mantadia reserve in the Eastern forest of Madagascar recorded the interaction of ants with Liparis chantaliae. Large endemic ants, Tetraponera grandidieri Forel (1891: 203), were seen to visit the flowers and eat or harvest a substance from the dark disk of the lip. The anther was partly dislodged by the ant but the pollen stayed in-situ, considering the shape and size of the insect, it could easily have effected pollination (Fig. 15D).
Pollination of orchids by ants is a rare event, they more commonly behave as nectar thieves but in some species they can act as additional pollinators (Claessens & Seifert 2018: 155).
Terrestrial or epiphytic, autotrophic or rarely holomycotropic herbs, usually with creeping rhizomes. Roots with a velamen. Stem cylindrical to swollen and cormous or pseudobulbous. Leaves one to several, thin-textured to fleshy, plicate or conduplicate, dorsiventrally flattened or iridiform, rarely terete, alternate or distichous, sheathing at base, articulated or not near base, rarely reduced to basal sheaths. Inflorescence terminal, flowers in spirals, whorls or distichously arranged, racemose or sub-umbellate, unbranched; peduncle and rachis cylindrical to markedly winged or angled, glabrous or hairy; floral bracts usually lanceolate, persistent. Flowers thin-textured, resupinate or apparently nonresupinate, often translucent. Sepals free or sometimes lateral sepals more or less fused, entire, usually spreading. Petals free, often smaller than the sepals, linear, often reflexed. Lip entire to lobed, recurved, lacking a spur, usually bearing a callus or ridges or mounds, often auriculate at base. Column often winged at apex, less frequently with lateral or basal wings, lacking a foot; anther terminal or sub dorsal, two-celled, motile, apical or dorsal, pollinia four in two pairs, often of unequal size, rarely two and deeply divided, laterally compressed, lacking a viscidium or with one or two minute viscidia; stigma entire, ventral; rostellum flap-like, often emarginated, seldom elongated. Ovary cylindrical, rarely papillose or winged. Capsule fusiform to suborbicular. (Adapted from Cribb in Pridgeon et al. 2005: 453) DISTRIBUTION. A tribe of 13 genera widely distributed in the Old and New World tropics and subtropics with a few representatives extending into North Temperate regions.
During field observations two of the authors (J. H. and J-M.H.) realised that there were two distinct species within the dwarf Madagascan Liparis with a single leaf and a distinct callus at the base of the lip. One is generally terrestrial with a relatively short leaf, an inflorescence invariably much longer than the leaf, without stolon and with a roundly lobed callus at the base of the lip. A second is generally epiphytic, with an inflorescence hardly longer than the leaf, with a distinct stolon and with a much larger conical callus at the base of the lip. Perrier's descriptions (1936: 248 -249) and(1939: 276) could apply to both these species but he was very clear that his Liparis monophylla had pale filiform stolons and he also used this characteristic to compare it with L. cespitosa, in addition he mentioned a prominent callus. His illustration (1939: 283) also shows long thin stolons. Although not obvious in the type material, the only other Liparis in the region reported (Perrier 1939: 283) to have a stolon is L. parva which has two leaves, flowers a third larger and a differently shaped anther and column (see Table 1).
Detailed analysis of the material indicated as types by Perrier clearly shows that they contain a mixture of two different species and this has been the cause of much confusion: all the plants on Decary 6222 and on Perrier 18548 have no stolons and habit and flowers that correspond clearly with Liparis scaposa which, during Perrier's time, was only recognised from Réunion. The sheet of François 4 (= Perrier 18548bis) has one flowering plant of L. scaposa but three plants with flower fragments and buds only but with clear stolons that correspond well with Perrier's description of L. monophylla. It is likely that Perrier assembled all plants with a single leaf that were not L. cespitosa as his new species, they may even have become mixed up, especially in specimen Perrier 18548bis.
It resembles Liparis bosseri, described below, but that species is much shorter and smaller, with a shorter inflorescence and has flowers with the anther with a spathulate beak (vs short and rounded). It differs from the mainland African L. bowkeri by the smaller plant and flowers with a different lip and column (Table 3). DISTRIBUTION. Probably endemic to Madagascar in Antananarivo, Fianarantsoa, Toamasina and Toliara provinces (Map 6). Szelengowicz & Tamon (2013: 358) cite and illustrate a plant assigned to this species from Réunion but because of the lack of vouchered material and the quality of the photographs it is impossible to ascertain if this identification is correct. One of the labels on the type material (P00094939) has 'Manongarivo' massif crossed out by Perrier and changed to 'Andringitra' massif. Manongarivo is in Northern Madagascar in the Tsaratanana mountain range. The Andringitra location fits with Perrier's collection numbers in that area and is therefore correct. SPECIMENS EXAMINED. MADAGASCAR. Andringitra massif, c. 1800m, Feb. 1922 lectotype P; Perrier 14396 (P00094940); Andringitra massif, Feb. 1922, c. 1800 (lectotype P); Perrier 14395 (P00094982) (isolectotype P); Andringitra massif, Feb. 1922, c. 1800; Ambatofinandrahana, Feb. 1919, Perrier 12420 (P00094985) (lectotype P); Perrier 12420 (P00094985) (isolectotype P); Mt Maromizaha near Analamazoatra, c. 1000 m, Feb. 1924, Perrier 16043 (holotype P);Central, Mt Ibity, c. 800 m, Feb. 1914, Perrier 7039ter (P); Ambohimanga forest, June 1915, Alleizette s.n. (P); Anjavidilana, Andringitra, Jan. 1950, Guillaumet 3720 (P, TAN); Andringitra, Jan. 1971, Guillaumet 3815 (TAN); High valley of the Menarakaka, E of Ihosy, 100 -800 m, , Humbert 28550 (P); Fianarantsoa, 39 km N of Ambositra, 1250-1350m, Jan. 1974, Croat 29475 (K, MO); Toliara, road from Ifarantsa to Enaniliha, 630 m, Feb. 1995, D. DuPuy et al. M869 (K); Kitsamby forest, near Soavinandriana, 1997, Hermans 3553 (K); Antananarivo, near Angavokely forestry station, Dec. 1997, Hermans 2107; along RN4 to Mahajunga, Dec. 1997 IUCN), and has this rare species has thus been preliminarily assessed as VU using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for smallscale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. Annual fires in the Ambatofinandrahana and Ibity areas have affected several colonies of the species. Its habitat is also affected by harvesting of wood for charcoal production in the Maromizaha reserve. FLOWERING TIME. December to May. ETYMOLOGY. Refers to the type locality of the Andringitra mountain range in SE Madagascar, now a National Park. Liparis henrici was named by Rudolf Schlechter for Henri Perrier de la Bâthie (1873-1958, L. latilabris refers to the broad lip, L. rectangularis presumably refers to the angles of the lip and L. verecunda probably refers to the diminutive nature of plant and flower. NOT ES. Schlechter's descriptions of Liparis andringitrana, L. henrici, L. latilabris and L. verecunda are just a few pages apart and could easily apply to all four species; strangely he does not compare them with each other but does so with other species. In the descriptions there are differences in the shape of the lip callus and curvature of the column which are never very clear in dried material. Liparis andringitrana, L. henrici and L. latilabris were collected by Perrier in the Andringitra mountains, presumably from almost the same spot with only one number difference in collecting code. Schlechter also illustrated all four species in 1932. Perrier in his revision of Liparis from Madagascar (1936: 242) made L. latilabris and L. verecunda synonyms of L. henrici, he did not explain the reasons. He first mentioned them under 'species not satisfactorily known' and then listed them as synonyms of L. henrici further in the text. In addition Schlechter was unaware that the name L. latilabris had previously been used by Rolfe (1903: 6) for an Indochina/Vietnam species.
On examination of herbarium material of Liparis andringitrana, L. henrici, L. latilabris, L. verecunda and L. rectangularis (Table 2) it becomes obvious that they consist of one, variable species. They all share the same plant habit, leaf-shape, lip, lip callus, column and  anther. There is a great variance in pseudobulb, leaf size and lip shape but there are a number of intermediate forms.
In plant and flower habit Liparis latilabris corresponds well with the typical form of L. andringitrana but the shape of the lip is puzzling: in his description Schlechter describes it as reniform and wider than long, this is reflected in his drawing and also in the type material but Perrier's drawing of the lip kept with the type shows a more typical oval form. On examining the type sheet of L. latilabris (Perrier 14394) the left hand plant has a lip that is broader than long whilst the other plants have a more typical oval lip of L. andringitrana. Liparis latilabris is therefore kept as a synonym.
Liparis verecunda is identical in plant habit, lip shape, column and anther to L. andringitrana and is also kept as a synonym.
Liparis rectangularis is known only from the holotype and consists of two plants collected in the Maromizaha forest in the east of Madagascar. The type is accompanied by a drawing by Perrier of the lip and anther. On dissecting the flowers it is obvious that his drawing somewhat exaggerates the size of the basal wings and the rectangular angles of the blade. This exaggeration is repeated in the original description and in his key (Perrier 1939: 284). It is a local variant of L. andringitrana with which it shares the same habit, flower size, lip and column, it is therefore considered a synonym (Table 2).
The lectotype material of Liparis henrici consists of generally bigger plants and a longer stem than those of the type of L. andringitrana (Table 2). The flowers however, are of a similar size and shape, the lip, lip callus, column and anther shape are identical to those of L. andringitrana. In his description and key Perrier (1939: 288) follows Schlechter's flawed description and singles out the flattened watery bulbs and obtuse rounded anther to differentiate L. andringitrana from the other species. The flattened bulbs are not uncommon in plants labelled as L. henrici, the pseudobulb shape is generally somewhat indistinct in dried material and their soft watery nature is likely to be environmental and has been observed in young pseudobulbs of other species, including L. densa. Perrier is adamant that the anther of L. andringitrana is rounded at the front (tout à fait arrondie en avant) but on dissecting the type material it is clear that the anther is shortly beaked as in the type of L. henrici. Lip and callus shape are also used as distinguishing characters between the different species but herbarium material has confirmed a great variety of lip shapes, even within a single collection. When dried, the shape of the callus remains as somewhat indistinct raised ridges only.
Extensive field observations by two of the authors (J. H. & J-M.H.) confirmed the great variability within the species: mature flowering plants within different colonies varied from six cm to more than 15 cm, some with distinct pseudobulbs and large leaves, others were more diminutive. Length of inflorescence was also found to be variable but the size and shape of the flowers was consistent. The callus of the lip was always prominent as was the anther beak. It was generally impossible to key-out or differentiate between the different populations in Andringitra and the Eastern forest around Andasibe. Based on these observations it was decided to treat Liparis andringitrana, L. henrici, L. latilabris, L. verecunda and L. rectangularis as one, variable species. Liparis andringitrana, based on Perrier 14396 (P00094939), was chosen as the most representative and complete. The type material of L. henrici is less representative of the typical form and has a few deformed flowers.
There are two herbarium sheets of the type of Liparis andringitrana (Perrier 14396); sheet P0094939 has the most representative plants and was therefore chosen as the lectotype, sheet P0094940 has a single plant and the label has 'Manongarivo' as the collecting locality on sheet P0094939, this is an error corrected by Perrier, it is chosen as an isolectotype. There are two herbarium sheets of the type of L. henrici (Perrier 14395); sheet P0094983 has the most representative plants and was therefore chosen as the lectotype, sheet P0094982 is an isolectotype. The same was done with the type material of Liparis verecunda. The type sheet Perrier 14394 of L. latilabris in P has three plants and there seems to be considerable variability in the flower, some flowers show the typical L. henrici lip but others have a somewhat deformed lip that is wider than long. ILLUSTRATIONS. Figs 3, 4;Schlechter (1932: T.51, T.53-4, 56); Hervouet (2018: 404).
Liparis bathiei resembles L. trulliformis, having the same habit, but differs in its smaller flowers, the relatively wider, rhombic, ecallose lip and the comparatively stout column. There are similarities in habit with L. lutea but the lip is bigger, the floral bracts larger and the anther cap has an obtuse rather than an acute beak. It also resembles L. bosseri, described below, in plant habit and shape of the anther but the rachis in L. bosseri is less dense, the flowers are twice the size, and the lip obovate (vs oval / lozengeshaped). It shares several characteristics with L. xanthina which may be the same species but there is not sufficient reliable material to support this. DISTRIBUTION. Endemic to Madagascar: Antananarivo province in the Ankaratra Massif in and near the town of Ambatolampy, recently also found a little further South in Fianarantsoa province (Map 7). Records from Réunion are misidentifications: photographs in Szelengowicz & Tamon (2013: 351) labelled as this species show an unrecognisable but different plant, inflorescence habit and flower, the description copies that of Perrier (1939: 291) for the species. Category EN: the extent of occurrence (EOO) of Liparis bathiei cannot be estimated since it is only known from two remaining subpopulations whereas its minimal area of occupancy (AOO) is estimated to be 8 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). This species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slashand-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. January to March. ETYMOLOGY. Named for Henri Perrier de la Bâthie (1873-1958 by Rudolf Schlechter. NOTES. Liparis bathiei was first found by Perrier in March 1921 and was sent to Schlechter the following month who described it three years later. Perrier made another collection in the same locality in February 1925 and gave them the same collecting number. In his description, Schlechter cited Perrier's specimen 13548 as the type with a collecting date of March 1921. Several sheets in G, K and P also have the number of Perrier 13548 but are from the later collection. As they come from the same locality as the holotype but were collected later, they can all be considered as topotypes.
Perrier in an overview of the geographical distribution of Liparis discussed the origin of this species (1936: 259): he first found plants in March 1921 on the immediate outskirts of the then small town of Ambatolampy, a few hundred metres South from the military barracks, at around 1500 m altitude, in an area that is now very barren, densely populated and cultivated. Further collections were made in  and duplicates deposited in G and K. According to Perrier the population was still there in 1932. He wrote that the plant grew in good numbers in the shade and in the humus of Acacia dealbata, introduced to Madagascar around 30 years before: "… There is no natural locality for this Liparis, remnants of original forest are more than 60 km away, and these remnants have been well explored by botanists. Finding this Liparis in these urban conditions under an introduced tree is interesting and not easy to explain except for originating from small seeds transported in the wind. It is also interesting that this species is new and has little resemblance to other Liparis found in the area (it is related to L. henrici but this species has only been found in Betsileo, i.e. more than 200 km S of Ambatolampy. It has reproduced with vigour in these conditions, very different from those where these species normally grow. Considering these circumstances it appears that this Liparis is a result of recent evolution, resulting from a seed transported by wind into a very different environment. This example of sudden mutation is especially interesting as it occurred on one of the main link roads of the Island, it is an interesting example of local evolution." Extensive searches by the first author in and around the town of Ambatolampy, especially near army barracks and underneath Acacia trees did not find plants of this Liparis. The species has now also been identified from the Mt Tsiafafavona in the Ankaratra Massif (Bosser 17711) which is only around 30 km from the type locality. A very small and vulnerable colony of plants was recently found c. 150 km S in Fianarantsoa province. ILLUSTRATION. Fig. 5;Schlechter (1932: T.52 Medium size, erect terrestrial, lithophyte or rarely epiphytic plant 14 -18 cm high, emerging directly from the base of the previous growth, roots filiform, more or less villous, c. 1 -2 mm diam. Pseudobulbs ovoid, narrowly elongate, entirely covered by 3 -5 brown membranous sheaths and by the sheath-like petiole of the lower leaves, with 3 -4 leaves, new growth emerging from the previous leafless pseudobulb which is almost completely disintegrated and often divaricate, 3 -4 × 0.5 -1 cm. Leaves erectly spreading, variable in size, blade ovate to broadly elliptic-acuminate, corrugate, with 7 pronounced ridges, narrowed into a short 15 -20 × 6 -8 mm sheath-like petiole amplexicaul to the pseudobulb, plicate, thin textured, pale green, overall 10 -16 × 5 -8.5 cm. Inflorescence erect, up to 30 cm long, c. 2 mm diam., on maturity far exceeding the leaves, carrying up to 20 flowers. Peduncle costate, about ½ the length of the inflorescence, with a few lanceolate sterile bracts cordate at the base. Rachis laxly racemose, up to 15 cm. Floral bracts shorter than the pedicellate ovary, spreading, ovate-lanceolate, attenuate at the tip, somewhat cordate at the base, green, 3.5 -9 × 2.2 -4.1 mm, becoming smaller towards the apex. Flowers small to medium in size, erectly spreading, overall up to 15 × 11 mm, entirely olive green, translucent, fading to ginger-orange. Pedicel and ovary porrect, cylindrical, somewhat winged, 8 -14 × 0.8 -1.2 mm. Dorsal sepal erect to incurved, ligulate, margins more or less recurved, 8.1 -9.9 × 0.7 -1.8 mm. Lateral sepals folded beneath the lip, elliptic, a little narrowed towards the base, 6.3 -8.1 × 2.5 -3.1 mm. Petals strongly curved backwards, linear, 7 -8.5 × 0.3 -0.8 mm. Lip shortly auriculate at the base, then elliptic-rectangular, sinuate and recurved at the anterior margin, with two distinct rounded horn-like calli at the base, 6.1 -8.2 × 5.1 -6.5 mm. Column curved towards the apex, with distinct rounded wings, 3 -5.5 × 1 -2 mm. Anther ovoid with a short acute beak at the tip, 1.1 -1.5 × 0.9 -1.1 mm. Pollinia ovoid c. 0.4 × 0.6 mm. RECOGNITION. Liparis bemarahensis is characterised by its ovoid pseudobulbs covered by sheaths, the deciduous previous growth, its thin-textured, ovate to elliptic leaves, long inflorescence with up to 20 flowers, the prominent floral bracts, small to medium-sized flowers with an elliptic-rectangular lip with two horn-like calli at the base, and the anther with a short acute beak. Liparis bemarahensis is similar to several of the softleaved deciduous Liparis species of the region. It is close to L. sambiranoensis but it is a smaller plant with smaller flowers, a shorter stem, and a rectangular lip (vs pandurate), with sinuate anterior margin (vs serrate) region (Table 4). It is vegetatively closest to L. perrieri but the flowers of the new species are larger, the rachis lax (vs sub-dense) and the anther distinctly beaked. There are also some similarities in habit to L. ochracea and L. ornithorrhynchos but the flowers are 1 3 smaller than those of L. ochracea, the lip elliptic-rectangular vs twice as long as broad in L. ochracea and rounded to broadly cordate in L. ornithorrhynchos and the anther with a much shorter beak than L. ornithorrynchos. It is also similar to L. andringitrana but the flowers are larger by 1 3 and more numerous, the lip elliptic-rectangular vs oval-flabellate and the lip with two distinct calli vs sub-angular. The distinct calli and lip-shape are considerably different from all the above. It is has a limited distribution in the Bemaraha area, at lower elevation. DISTRIBUTION. Madagascar: Mahajanga province only, restricted to the Bemaraha plateau and 'Tsingy' limestone karst (Map 8). Near Ambodiriana, E of Antsalova, 100 -150 m, Jan. 1960, Léandri & Saboureau 2687 (holotype P (P01778638), isotype P (P01778635)); towards Ambodiriana, E of Antsalova, 100 -150, Dec. 1952, Léandri et al. 2171; Antsingy of Antsalavo, Bemaraha, Jan. 1975  HABITAT. Deciduous forest on limestone Tsingy, chalk rock, dense low deciduous forest with a canopy of various heights between c. 10 -20 m, limestone boulders. Altitude: 100 -200 m CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis bemarahensis cannot be estimated since it is only known from two remaining subpopulations whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). This species is only known from two subpopulations representing two locations (sensu IUCN), and has thus been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. December to February. ETYMOLOGY. Referring to the Bemaraha area, a nature reserve and UNESCO World Heritage site, where this species was first found. NOTES Since then a number of other collections from the same locality have been mis-or unidentified in herbaria. There are two sheets of the Léandri 1960 collection in P, in both the plants are somewhat damaged but the flowers are intact and therefore make suitable type material. ILLUSTRATION. Fig. 6. Liparis bernieri is similar to L. danguyana and L. listeroides which are closely allied but distinct morphologically and their distribution is disjunct ( Table 5). The leaf of L. danguyana is thinner in texture than those of L. bernieri and L. listeroides which are more leathery with indistinct veins, the leaf shape is ovate, sessile vs more broadly oval, sub-sessile. The flowers of L. listeroides are generally smaller than the other two. The lip of L. danguyana has distinct basal lobes, that of L. bernieri does not and is almost pyriform, it has two small but distinct calli at the base (in L. bernieri the callus is indistinct and consists of a crescent-like cushion), a front margin that is slightly attenuate at the front margin vs emarginate in L. bernieri. The anther cap of L. bernieri is rounded at the front with a short angular beak, that of L. listeroides is longer and obtusely beaked. DISTRIBUTION. Endemic to Réunion: in isolated localities in the interior of the Island (Map 9). SPECIMENS EXAMINED. RÉUNION HABITAT. Wet forest, often on humus and mosscovered surfaces, in shade. Altitude: 800 -1600 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis bernieri cannot be estimated since it is only known from two remaining subpopulations whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). With only two known subpopulations representing two locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by grazing and anthropogenic fires, resulting in habitat reduction and habitat quality reduction.
Previously assessed and published in the IUCN Red data listing as EN (Picot 2013: 17). FLOWERING TIME. January to March. ETYMOLOGY. Dedicated to Dr Charles Bernier (1802-1858, director of the Jardin Botanique de La Réunion. NOTES. The species was first mentioned as 'Liparis bernieri Frappier' in his listing of the plants of Réunion (Frappier 1880: 15). It was not formally described until 1895 in Cordemoy's Flore, the orchid part being based on Frappier's manuscripts. The description indicated that the author knew the plants in the wild and had also found herbarium material in the Réunion Museum where it was labelled as Liparis elegans Bernier, a name that had already been used five years earlier by Lindley for an Asian species. This group of unmounted plants, presumably collected by Bernier, found its way into the Cordemoy herbarium in Marseille where it is accompanied by labels of its origin and the same manuscript text used in Frappier's description. This is most likely the original material seen by the author; it has corresponding labelling and is close to the original description and is therefore without doubt the holotype of the species and has priority, according to ICN article 9.19, over the lectotype chosen by Margonska (2009: 92) as outlined below. Finet (1909: 98, pl.1) included a description and illustration of the species and mentioned that the plant was found mixed with Liparis purpurascens (now L. salassia), citing both Commerson and Boivin collections.
Margonska may well have been unaware of Cordemoy's Marseille herbarium material and lectotypified the species in error using Boivin's specimens kept in P. The Cordemoy material from MARS had been on loan for many years to P and not directly accessible, lately it has been on loan to the University of Réunion (REU). In the same paper Margonska (2009: 92) also considered Liparis danguyana to be synonymous with L. bernieri citing their similar characteristics. As explained above, there is a strong case for keeping these species separate.
Anther ovate-obtuse c. 1 × 1 mm. Pollinia c. 0.5 mm. RECOGNITION. Liparis bicornis is distinguished by having a single flowering growth with the previous year's growth disintegrated, erect leaves, a tall inflorescence with many small flowers, with an obcuneate emarginated lip with two small but distinct calli at the base and an anther with a rounded margin.
In habit it is similar to Liparis densa but that differs in its lip shape (ovate-obcuneate vs obovate-oblong) and lack of distinct calli at the base of its lip. Its lip is similar to that of L. perrieri but the anterior lip margin in that species is dentate rather than rounded. DISTRIBUTION. Endemic to Madagascar: mainly in the highlands of Antananarivo province with a few colonies in the Eastern forests (Map 10).  HABITAT. Wet marshy slopes, rocky outcrops. Altitude: 1300 -1500 m CONSERVATION STATUS. Category VU: the extent of occurrence (EOO) of Liparis bicornis is estimated to be 32,601 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 28 km 2 (which falls within the limits for Endangered status under the criterion B2). With seven known subpopulations representing seven locations (sensu IUCN), this species has been preliminarily assessed as VU using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. The Ankatso habitat was destroyed for building and Eucalyptus plantations. The habitats near Ambatolaona and Lakato were burnt for cultivation. FLOWERING TIME. January to June. ETYMOLOGY. The name refers to the two distinct horns at the base of the lip. NOTES. The species was first found on wet mountain slopes by the German collector Johann Hildebrandt in 1881, a few months before he died in Madagascar. The type locality of the 'Imerina' area normally refers to the Madagascan Highlands around the capital Antananarivo. Henry Ridley acquired some of Hildebrandt's herbarium and described the new species four years later and is clearly based on the BM specimen. In line with much of Hildebrandt's herbarium the collections were widely sold or distributed around Europe, in this instance G, K, M, P, W and WU. A detailed drawing of the flowers by Ridley is in the orchid herbarium at K (Fig. 8). ILLUSTRATIONS. Fig. 8;Perrier (1939: 283 Very small, squat terrestrial or lithophytic plant up to 6 cm high, on a very short to almost non-existent rhizome, roots wiry. Pseudobulbs turbinate, whitish, 10 -18 × 7 -15 mm, covered with 2 -3 thin membranous sheaths, carrying 2 leaves, very rarely only one, the previous year's pseudobulbs without leaves. Leaves spreading, ovate, apex acute, base cuneate to obtuse, thin, strongly veined, 4 -6.2 × 2.2 -4.3 cm. Inflorescence erect, much taller than the leaves, up to 13 cm, c. 1.5 mm diam., with 2 -10 flowers, opening in succession. Peduncle about ½ the length of the inflorescence, with 3 -4 prominent peduncle sheaths, 6 -9 × 2 -3 mm. Rachis loosely racemose. Floral bracts lanceolate, decreasing in size towards the apex of the rachis, 3 -6 × 2 -2.9 mm. Flowers medium in size, 10 -12 × 6 -7 mm, greenish, becoming yellow with age. Pedicel and ovary slender, grooved, 7 -12 × 0.7 -1.6 mm. Dorsal sepal erect, recurved towards the apex, narrowly lanceolate, 6.1 -9.2 × 1 -1.9 mm. Lateral sepals broadly falcate, obtuse, 6.1 -8.9 × 2.8 -3.2 mm. Petals linear, obtuse, 6.3 -9 × 0.2 -0.3 mm. Lip obovate, acuminate at the tip, anterior margin a little undulate, longly auriculate at the base, thickened ridges at the base becoming a canaliculate longitudinal ridge along the disk, 6.1 -7.5 × 4.2 -6.1 mm. Column almost straight, wings longly rounded, 3.
Anther with a distinct spathulate beak, c. 0.8 mm long, rounded at the apex and narrowed towards the base, overall c. 1.7 × 0.9 mm. Pollinia pyriform c. 0.9 × 0.6 mm. RECOGNITION. A very squat plant with broad, spreading leaves, a relatively long inflorescence and proportionately large flowers. Its lip is acuminate at the tip, longly auriculate at the base and lacks a distinct callus but has a thickened ridge formed of the three raised central veins.
Liparis bosseri has a number of characteristics overlapping with other species but none in the same combination. It is closest to L. trulliformis but the lip of L. bosseri is obovate (vs ovate-trullate), lacks a distinct callus at the base of the lip and has an anther with a spathulate (vs a sharp beak). In habit, flower size and lip shape it resembles L. andringitrana but L. bosseri is much squatter and smaller, the inflorescence shorter and the anther has a spathulate beak (vs short and rounded). It also resembles L. bathiei in habit and shape of the anther but the rachis of L. bosseri is less dense, the flowers are about double the size, the lip obovate (vs oval-lozenge-shaped). The habit and flowers of L. nephrocardia are similar in size but the inflorescence of L. bosseri far surpasses the leaves (vs about the same length) and lacks the distinct rounded callus of the lip. It shares the spathulate beak of the anther with L. zaratananae but the plant of L. bosseri is 1 3 the size, without a distinct stem and also has smaller flowers and an obsolete callus vs ridges at the base of the lip. DISTRIBUTION. Endemic to Madagascar, South-Central, Toliara and Fianarantsoa provinces (Map 11). SPECIMENS EXAMINED. MADAGASCAR. Near Betroka, Feb. 1963, Bosser 17910 (holotype P); Ihorombe Region, Ihosy  (AOO) is estimated to be 8 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). With only two known subpopulations representing two locations (sensu IUCN), this rare species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. February and March. ETYMOLOGY. Named for Jean Bosser (1922Bosser ( -2013, a great authority on the Flora of Madagascar and the Mascarenes. ILLUSTRATIONS. Fig. 9. Liparis bulbophylloides H. Perrier (1936: 242;1939: 263); Hermans et al. (2007: 216); Cribb & Hermans (2009: 145). Type: Madagascar, Toamasina, confl. Onive & Mangoro, 1925, Perrier 16998 (holotype P).
Very small epiphytic plant up to 2.5 cm high, caespitose, rhizomatous epiphyte up to 10 -15 cm long, the growths 5 -25 mm apart, rhizome 1 -3 mm diam. covered in brownish papery sheaths, roots 0.3 -1 mm diam., somewhat hairy, the tips glabrous. Pseudobulbs, small, oval, 8 -16 mm long, c. 5 mm diam., base covered by a membranous sheath 5 × 4 mm, carrying 2 leaves at the apex. Leaves ovate subsessile, rounded at the base and shortly acute at the apex, 10 -18 × 8 -15 mm. Inflorescence erect from the centre of the new growth, 3 -10 cm tall carrying 3 -10 flowers and buds. Peduncle with several bract-like sheaths. Rachis somewhat twisted, flowers 4 -12 mm apart. Floral bracts green, with rounded wings, acute at the tip, single-veined and about 1 3 the length of the pedicellate ovary, 4 -6 × 3 -4 mm. Flowers small, overall c. 8 × 5 mm, opening in succession, all segments recurved at the margins, pale yellow-green with the lip darker yellow-green, column white, pollinia yellow, flowers becoming brownish-orange with age, flower buds bulging at the base. Pedicel and ovary ridged, 5.5 -8 × 0.6 -1.2 mm. Dorsal sepal folded backwards, lanceolate, slightly acute, 4.2 -6.8 × 1 -1.6 mm. Lateral sepals folded underneath the lip, oblong to oval, obtuse 3.1 -5 × 1.8 -2.5 mm. Petals linear, single veined, 4 -6 × 0.4 -0.8 mm. Lip suborbicular, trisinuate, narrowed and slightly auriculate at the base and incurved from the middle, the blade opaque with a few obscure veins, a distinct (0.5 mm) rounded cylindrical or bilobed callus at the base, 3 -4.1 × 3.1 -4 mm. Column arched, the wings strongly expanded at the side of the stigmatic surface into prominent angular membranous lobes, then decurrent lower down, rostellum with a very small central tooth, 3 -3.8 × 1 mm. Anther extended at the front by a rounded semi-cylindrical beak almost the same size as the cap, 1 -1.3 × 0.5 -0.7 mm. Pollinia four, fused together in two pairs, oblong, 0.3 -0.4 × 0.2 mm. RECOGNITION. The species is one of the few truly epiphytic Liparis in the region. In habit it resembles species of Bulbophyllum Thouars (1822: t.3) with two small oval apical leaves and a small rounded pseudobulb on a long rhizome. The flowers are small, the lip is suborbicular with a rounded callus at the base, the column has obtuse transparent wings, the anther cap has a semi-cylindrical obtuse beak almost as large as the cap itself; somewhat resembling a baseball cap.
Liparis ambohimangana is similar but has a true stolon rather than a rhizome, a single leaf vs two and the anther is not beaked. DISTRIBUTION HABITAT. Exclusively as an epiphyte in humid evergreen forest, amongst mosses on tree branches. Altitude: 600 -1100 m. CONSERVATION STATUS. Category VU: the extent of occurrence (EOO) of Liparis bulbophylloides is estimated to be 71,015 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 44 km 2 (which falls within the limits for Endangered status under the criterion B2). With 11 known subpopulations of this species representing 11 locations (sensu IUCN), L. bulbophylloides has been preliminarily assessed as VU using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. January to March. ETYMOLOGY. Refers to the shape of the pseudobulbs and plant habit similar to some species of the orchid genus Bulbophyllum. NOTES. The species was first described by Perrier from a collection made by Henri Humbert in Eastern Madagascar. It has since been found in a wide area along the Eastern mountain range of the Island. ILLUSTRATIONS. Figs 10, 11;  Medium to large erect terrestrial or lithophytic plant, 14 -33 cm high, rhizome short, ascending, roots wiry, more or less pilose, c. 1 mm diam. Pseudobulbs slender, stem-like, a little inflated at the base, up to 25 cm × 6 -15 mm, partly covered by 4 -5 thin amplexicaul sheaths, partly overlapping, up to 5 cm long, with 2 -3 (rarely 4) apical sub-petiolate leaves. Leaves spreading, asymmetric, lanceolate to elliptical-lanceolate, acuminate, undulate, the veins obliquely convergent, more or less glossy, contracted into a 10 -25 mm petiole at the base, overall 8 -10 × 4 -5.5 cm, the third leaf usually smaller, dark green. Inflorescence erect, a little longer than the leaves, carrying up to 15 flowers but usually fewer, up to 11 cm. Peduncle with a few linearlanceolate peduncle sheaths becoming sterile bracts, 7 -10 mm long. Rachis loosely racemose in the upper half of the inflorescence, up to 7 cm long. Floral bracts spreading, linear-lanceolate, acuminate, the lower ones considerably longer than apical ones, up to 13 × 3 mm, green. Flowers medium in size, up to 15 × 10 mm, ovary green, sepals and petals yellowish-brown, Fig lip dark brown to ochre, darker towards the middle becoming a dark orange-red on the disk and base of column, column white becoming brown towards the base; the whole flower turning more orange with age.
The species shares a similar habit, including the somewhat asymmetric leaves, its brownish-yellow flower colour and lip shape with Liparis rivalis from Madagascar but plants of L. caulescens are a little shorter, the spike habit is laxly racemose (vs densely racemose), fewer and with larger flowers, the lip with a more distinct callus, a larger, differently shaped column with angular vs rounded wings (Table 6). They may well have evolved from a common ancestor. It is somewhat similar in habit to Liparis gracilipes and L. danguyana from Madagascar but the leaves are considerably longer and asymmetric lanceolate vs elliptic-oval, while its flowers are at least 1 3 larger and the lip broadly obovate vs suborbicular. DISTRIBUTION. Réunion: all the cantons, mainly Saint-Denis (le Brûlé), Saint-Pierre (Tampon), Saint-Louis (Cilaos) (Map 13). Humblot s.n. P00209845 (P) from the Comoro Islands, identified as this species, is possibly a misidentification and there appear to be two different species on the sheet both bearing only seed-pods. However, a sterile plant collected on    '. From the above it seems that Richard intended to describe the species as Liparis vaginata but never did so. The Richard specimen at K-Lindl. was traced on paper by Reichenbach f. (now in W) and the tracing is annotated 'Liparis vaginata A. Rich. sp. nov' but additionally as 'Lip. caulescens Boiv. Isle Bourbon' and also labelled as 'L. longa Rchb.f cf. Linnaea". Liparis longa was described by Reichenbach in Linnaea (1877: 98) at the end of a lengthy series of novelties, most with a reference to type material except for L. longa which was the last in the series. The description is fairly detailed and mentioned a plant with a long stem, sub-opposite leaves and a lip with two teeth at the base of the lip. It is likely that Reichenbach was referring to the Richard material and decided to apply his own epithet and not the one suggested by Richard. Liparis longa is an earlier name for the species but under the rules of nomenclature (ICN article 7.9) Reichenbach's name cannot be accepted as it lacks any type material. In 1887, Ridley included L. longa in his monograph but added no further information.
Liparis caulescens first appeared in Frappier's listing of the plants of Réunion (1880: 15) as 'Liparis caulescens Frappier (ex. Boivin)'. It was finally validly described in 1895 by Cordemoy (1895: 186). Frappier did not specify a holotype in his description apart from indicating that it was based on a name given by Boivin on a herbarium specimen: 'Frapp, ex Boiv. hb. Mus. Réun. '. There are a number of herbarium sheets from the 1847 -1852 Oise expedition to Réunion in P and in K: some give Les Hauts du Boucan Launay as a more specific locality, most have the collecting number '1035'. The most representative sheet of original material used by Frappier is P00112460 and was designated here as lectotype. Sheets P00112455, P00112456 and K000718243 are isolectotypes. The P specimen P00112461 has no collecting number and is treated as a lectoparatype. A specimen in the Paris  Stichorkis cespitosa (Lam.) Thouars (Thouars 1809: 318) referring to Epidendrum cespitosum of Lam. Malaxis caespitosa (Lam.) Thouars (Thouars 1822: t.90 (Schlechter 1911b: 209).
Column arcuate, slender, the edges a little expanded towards the apex, 1.1 -1.6 × c. 0.5 mm. Anther small, obtuse, slightly expanded at the front, 0.4 -0.6 mm diam. Pollinia elliptic, 0.2 mm. Seed capsule globular, ovate, c. 2 mm diam. with 3 strong ridges. RECOGNITION. A small plant with rounded-ovoid pseudobulbs, bearing a single leaf, an inflorescence generally a little longer than the leaves, with up to 20 flowers, with 3 mm long sepals, an oblong and recurved lip, lightly denticulate at the anterior margin, and lacking a distinct callus, an unwinged column only slightly expanded towards the tip and an anther obtusely expanded at the front. The length of the floral bracts is variable, sometimes exceeding the length of the flower. Seed capsules are persistent on the rachis. It shares the small pseudobulb and leaf-shape with Liparis ambohimangana but is very different in lacking a stolon while its flower, lip and anther are much smaller and a different shape with a suborbicular vs oblong-obtuse lip and a distinct callus. It is similar in habit to L. scaposa but has considerably smaller flowers and its lip callus is less apparent ( Table 1).
high Bara valley, Feb. 1971, T. Cadet 3058 (P); Dos d'Âne, Feb. 1971 HABITAT. Moss forest, secondary forest, humid forest. Altitude: 0 -1400 m. CONSERVATION STATUS. Category LC: the extent of occurrence (EOO) of Liparis cespitosa is estimated to be 700,161 km 2 (far exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 64 km 2 (which falls within the limits for Endangered status under the criterion B2). With 14 known subpopulations representing 15 locations (sensu IUCN), this widespread species has been preliminarily assessed as LC using the green listing method. In Madagascar, this species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction Previously assessed and published in the IUCN Red data listing for Réunion as CR (Picot 2013: 11) FLOWERING TIME. January to April. ETYMOLOGY. Referring to the tufted habit of the plant, as first described by Lamarck. NOTES. Lamarck (1783: 187) first described Liparis cespitosa as Epidendrum cespitosum based upon a Bourbon [Réunion] specimen from de Jussieu's herbarium: he refers to it as 'Angrec en gazon' and likens its habit to E. ophioglossoides Jacq. (Jacquin 1760: 29) (= Stelis ophioglossoides (Jacq.) Sw. (Swartz 1800: 239)). The holotype specimen, as used by the author, is in the Lamarck Herbarium in P (P00382694), it consists of a fruiting plant. Later labelling on the sheet refers to Stelis and E. ophioglossioides and also to 'Martinica' obviously referring to the Antillean Stelis species in error. In addition, the de Jussieu herbarium has a sheet with five fruiting plants and one gap with the pin-holes of a previous mounting still in place, it again is annotated Stelis and E. ophioglossioides and also as Malaxis caespitosa Thouars. The Lamarck holotype specimen was without much doubt removed from this de Jussieu sheet which is, therefore, considered to be an isotype. It is likely that the original material came from Commerson but it is inconclusive whether the 'Comm' annotation on the Jussieu sheet refers to Commerson or to its communication. It is also interesting to note that a specimen sheet of unknown origin in the de Candolle Herbarium in G, has a mixture of two plants labelled as E. ophioglossiodes with one plant of L. cespitosa and another of a plant similar to S. ophioglossoides.
In 1809 Du Petit-Thouars referred to Lamarck's description in error under his new genus 'Stichorkis' as Angraecum caespitosum. In 1822 he referred to it both in the text and in the illustration as Malaxis caespitosa; there is no reason to believe he would not be referring to Lamarck's Epidendrum cespitosum. Lindley placed it in the genus Liparis in 1824. Margonska (in Szlachetko et al. 2008: 35) rejected Rasmussen's (1979: 392) interpretation of Du Petit Thouars' work (1809Thouars' work ( & 1822 and re-typified and created Stichorkis cespitosa (Lam.) Thouars ex Marg.
The nomenclature system allows a number of interpretations under the nomenclatural code (McNeill et al. 2012& Turland et al. 2018) but this is superfluous: Liparis cespitosa is clearly based on Lamarck's ex de Jussieu's original material as used by the author and not on any of Du Petit-Thouars' material.
Both the epithets caespitosa and cespitosa have been used in the literature, under the rules of nomenclature and as discussed by Rasmussen (Rasmussen 1979: 392) cespitosa is correct, keeping Lamarck's orthography.
This is one of the few Liparis in the Madagascar area with small bulbous pseudobulbs and a single leaf. It was first recorded in the area from the Mascarenes but has since also been recognised from Madagascar where it seems to be fairly widespread. It is also one of the few species with a much wider distribution in Africa and in Asia. Percival (1965: 11), mentioned that it is cleistogamous in the wet and misty rainforest but chasmogamous in the drier air outside of that environment. ILLUSTRATIONS.  Large terrestrial or rarely epiphytic plant up to 26 cm tall, rhizome scrambling, sometimes new growths emerge from the stem of previous growth, roots glabrous, white, c. 2 mm diam. Pseudobulbs stem-like, round, up to 23 cm long, 8 -15 mm diam., with 3 -6 white sheaths along its length not much overlapping, attenuate at the apex, white-grey, strongly veined, dotted with ochre-brown, carrying two (rarely 3) leaves at the apex. Leaves broadly ovate, obtuse or shortly acute at the apex, undulate, crenulate along the margins, subsessile or shortly petiolate, pale to dark green, often a little paler underneath, 5 -8.5 × 2 -4.5 cm. Inflorescence erect, 8 -15 cm long, carrying up to 14 flowers but generally fewer. Peduncle about the same length as the rachis, with one or two peduncle sheaths enveloping the base and one or two higher up, c. 2 mm diam. Rachis laxly 2 -14-flowered. Floral bracts lanceolate, the base somewhat cordate, 5 -15 × 1.5 -3 mm. Flowers large, up to 41 × 23 mm, pale green to viridian, the lip dark olive green with a darker longitudinal ridge, column and ovary white, anther green, pollen yellow, scented. Pedicel and ovary somewhat ridged, thickened at both ends, 11 -16 × 1 -1.2 mm. Dorsal sepal erect, linear, obtuse, margins recurved, 13 -23 × 1.1 -2 mm. Lateral sepals folded below the lip, falcate, broadly lanceolate, 13 -19 × 3.5 -7.1 mm. Petals narrow linear, curved downwards, the margin incurved, 13 -19 × 0.3 -0.8 mm. Lip broadly oval-lanceolate, the base with short rounded wings and two elongate lamellar-triangular calli, generally with a furrow of smaller, irregularly rounded calli in-between, anterior lobe almost triangular with dentate margins, 11.5 -15 × 5.8 -8 mm. Column slender, slightly arching, wings narrow and obtuse, 4.5 -7 × 1 - Anther with a cuspidate to linear beak, 1.1 -1.8 × 0.8 -1.3 mm. Pollinia ovoid, c. 0.6 × 0.5 mm. RECOGNITION. Liparis chantaliae is easily recognised by its tall pseudobulbs, circular in cross-section and partly covered by papery sheaths, with two broadly oval, corrugate leaves at the top. The large flowers have a characteristic broadly oval-lanceolate lip, with short rounded wings and two elongate lamellar-triangular calli at the base, and an almost triangular anterior lobe with dentate margins. The anther has a cuspidate to linear beak.
Liparis chantaliae is similar to L. gracilipes but the leaves are much thinner in texture and broadly oval vs elliptic, the lip broadly oval-lanceolate vs suborbicular and the callus bi-lamellar vs tridentate. There are some superficial similarities with L. zaratananae but the pseudobulbs are shorter, less leafy, the lip is ovallanceolate with the anterior margin angular vs broadly oval and rounded, the anther has a shorter, less spathulate beak. It shares the tall habit, somewhat corrugate leaves, lax inflorescence and large flowers with L. longicaulis with which it has frequently been confused in herbaria and in field-observations  (Table 7), but the latter differs by the angular stem which is completely covered by overlapping papery sheaths (vs partly covered), generally longer inflorescence, larger flowers, a lip with an attenuate anterior lobe (vs rounded), a distinct bilobed callus on the lip (vs rounded swelling) and a more acutely beaked anther. The two species have a wide and overlapping distribution but L. chantaliae has been recorded from Toliara province, where L. longicaulis is absent. The species overlap in distribution in Andasibe (Toamasina prov.) and Ranomafana (Fianarantsoa prov.), they grow at similar elevations in wet evergreen forest, flower at similar times and both are scented. DISTRIBUTION  Schlechter, in his description, wrote that this species is similar to Liparis epiphytica Schltr. (Schlechter 1905b), from Western Africa: in several characteristics the plant and habit are similar although the leaves in L. epiphytica are blunter, the flowers of L. cladophylax are much smaller (about half the size), the lip basal formation is also different ('M'-shaped vs rectangular). There are some similarities to L. cespitosa but that has a single leaf, and there are also differences in the lip which is rounded in L. cladophylax and more expanded and dentate in L. cespitosa. It is also close to L. scaposa but the flowers are smaller, the column has smaller wings and a smaller callus on the lip. As the species is only known from limited material, it is difficult to assess its identity with any certainty. The original description calls for a dense spike of 15 -20 flowers but the herbarium material bears no resemblance to this, this characteristic would make it very different from similar small species; with this conflict between type material and the description the species will have to remain somewhat ambiguous. DISTRIBUTION  HABITAT. Epiphyte in humid evergreen forest. Altitude: c. 800 m CONSERVATION STATUS. Category CR: the extent of occurrence (EOO) of Liparis cladophylax cannot be estimated since it is only known from one subpopulation whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2).
The only known subpopulation represents one location (sensu IUCN). This species has thus been preliminarily assessed as CR using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slashand-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. October. ETYMOLOGY. Possibly referring to the dominant leaves or to the epiphytic habit of the species. NOTES. The holotype, collected as Perrier 139, was then changed to Perrier 11360 and is lodged in Paris; this is sometimes cited as an isotype but it is clearly marked as Schlechter's type as used for the protologue and should therefore be considered the holotype of the species, especially as any other of Schlechter's material would be likely to have been destroyed in Berlin. ILLUSTRATIONS. Schlechter (1932: t. 52

RECOGNITION.
A medium-sized plant with a linear rachis and fleshy pseudobulbs and leaf. The leaf is flat, single, broadly ovate and clearly cordate at the base, the mid-vein is indistinct, the floral bracts cordate, and the flowers large for the genus, with an ovate to subrectangular lip with an indistinct callus, a column with rounded wings and an anther cap with a long sub-acute beak. It is long and successively flowering.
Liparis clarae is similar to L. warpurii but differs by the distinct habit with a short linear rachis (a longer scrambling habit in L. warpurii where new growths emerge from the lower nodes of the previous pseudobulb), the wider, flatter, heart-shaped leaf without a distinct central vein, cordate floral bracts and larger and more numerous flowers. As shown in Table 8, it has similarities with several other species, including L. magnifica, described below but L. clareae is around three times the size in both plant and flower size and its callus is indistinct (vs bidentate). It is related to L. laurentii, described below, but the plant of L. laurentii is smaller, the leaf has a more distinct central vein and the base is rounded (vs cordate), it has 1 -5 flowers (vs 6 -12), the transversely elliptic or reniform lip is much wider (vs ovate), the margins are not serrate, the anterior margin is emarginate (vs entire), the basal calli consisting of horn-like calli (vs slight swellings) and the column wings are elongate (vs short and obtuse). It is also close to L. superclareae, described below, with which it shares its general habit and lip shape but L. superclareae has much longer pseudobulbs, the leaves of the previous growth disintegrate before the new ones develop (vs persistent), the leaves are lanceolate (vs ovate) and at least 2.5 -3 times longer than wide (vs 2 times), the floral bracts are less cordate at the base, the flowers and all its segments much larger, the column shorter with broad long wings (vs short-auriculate) and the anther more sharply beaked (Table 8). DISTRIBUTION Hermans et al. 2007: 289) was described by Perrier in 1936 as a tentative variety of the species, it is only known from the type specimen which corresponds well with the characteristics of L. warpurii and is included and discussed under this species. Perrier, in his description, noted that single pollinia were found in each of the two anther chambers in the type, this peculiarity seems to vary between specimens with some showing the normal two pollinia per chamber. ILLUSTRATIONS. Fig. 16;Hermans et al. (2007: pl.45); Cribb & Hermans (2009: 147); Hervouet (2018: 406).
Liparis danguyana H. Perrier (1936: 248; Medium-sized, erect terrestrial or epiphytic plant, 11 -18 cm, rhizome ascending, c. 5 mm diam., roots fibrous, wiry. Pseudobulbs 1 -2 cm apart, stem-like, 7 -12 cm × 1.5 -5 mm diam., carrying two leaves at the apex, enveloped by 5 -8 scarious sheaths. Leaves dark green, broadly oval, subsessile, shortly acute at the apex, rounded at the base, petiole c. 8 × 1.5 mm, blade strongly veined, 2.4 -4.5 × 1.8 -3 cm. Inflorescence up to 7 cm long. Peduncle with 3 -4 bract-like sheaths. Rachis erect, subdensely 4 -12-flowered. Floral bracts acute, about 1 3 of the length of the pedicellate ovary, 4 × 2.1 mm. Flowers small, c. 6 × 8 mm, green. Pedicel and ovary 8 -9 × 0.8 -1.8 mm, faintly winged. Dorsal sepal lanceolate, not auriculate at the base, 7 -10 × 1.2 -2 mm. Lateral sepals semi-oval, rolled below the lip, 5.5 -7 × 3 -3.5 mm. Petals linear, 3-veined at the base, 7.5 -10 × 0.8 -1.2 mm. Lip ovate to orbicular, obscurely veined, the basal wings obtuse and divaricate, slightly attenuate at the anterior margin, with 2 small but distinct rounded calli at the base, 6 -6.5 × 4 -5.5 mm. Column curved at the apex, 4 -4 × 1 mm, wings thin and small. Anther suborbicular, very rounded at the front, 1 -2 pollinia per chamber, 1 × 0.8 mm. Pollinia ovate, c. 0.2 mm diam. Margonska (2009: 97) mistakenly considered this species to be conspecific with Liparis bernieri. As shown in Table 5, there are a number of superficially similar species with an equally long fleshy stem and two leaves from the area, including: L. bernieri from Réunion and L. danguyana and L. listeroides from Madagascar but there are substantial differences: The leaves of L. danguyana are thinner in texture and have a more cordate base than those of L. bernieri and L. listeroides which are more leathery and have indistinct veins, the leaf shape of L. listeroides is oval petiolate vs broadly oval and sessile. The flowers of L. listeroides are generally smaller than the other two. The lip of L. danguyana has two small but distinct calli at the base, that of L. bernieri is cushion-like, those of L. listeroides are long and tooth-like, the anterior margin of L. danguyana is apiculate at the anterior margin, the others have a more or less denticulatesinuate margin. The anther of L. danguyana is rounded at the front and suborbicular, that of L. listeroides is slightly beaked. Considering these differences and specific geographical distribution it is correct to recognise all three species (Table 5). DISTRIBUTION. Endemic to Madagascar, until recently known from the type locality of Angavokely in Antananarivo province only. It has now also been recorded from Ambondrombe in Fianarantsoa prov. Category CR: the extent of occurrence (EOO) of Liparis danguyana cannot be estimated since it is only known from one remaining subpopulation whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). Liparis danguyana is only known from one subpopulation representing one location (sensu IUCN), this rare species has thus been preliminarily assessed as CR using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slashand-burn farming), resulting in habitat reduction and habitat quality reduction. The type locality now has only remnants of forest and little surrounding habitat. FLOWERING TIME. February to March. ETYMOLOGY Small, compact, erect terrestrial or lithophytic plant, 10 -15 cm tall, rhizome very short, roots wiry, glabrous. Pseudobulbs oblong, almost completely submerged in soil and humus, 2-edged, compressed, 2 -3 × 1.2 -1.4 cm, watery, turning into pulp under finger pressure, carrying 2 -5 leaves, the base enveloping the pseudobulbs. Leaves erectly to suberectly spreading, oblong-ligulate, acute or somewhat obtuse, a little fleshy, 6 -8 cm long, 1.2 -2.5 cm wide. Inflorescence up to 11 cm long, with none or one thin sheath towards the base, carrying up to 15 flowers. Peduncle ridged, up to 6 cm, hardly longer than the leaf sheaths. Rachis densely flowered, cylindrical, much longer than the peduncle, up to 8 cm long, 1.7 mm diam. Floral bracts narrowly lanceolate, acuminate, the lower ones as long as the ovary or longer, the upper ones shorter, up to 10 × 2.1 mm. Flowers small, 10 × 5 mm, uniformly pale yellow to yellow. Pedicel and ovary rounded, slightly ridged, 8 × 0.8 mm. Dorsal sepal erect, ligulate-lanceolate, obtuse, 6.5 -8 × 1 -1.3 mm. Lateral sepals subfalcate, oblong, obtuse, spreading, curled backwards, 4.8 -5 × 2.2 -2.7 mm. Petals erect or semierect, linear, somewhat obtuse, 6.5 -8 × 0.8 -1 mm. Lip distinctly geniculate, recurved, obovate-oblong, very obtuse, margins sub-crenulate, basal margins minutely triangular-auriculate, creating a deep gutter at the base, without a callus on the disk and base, 4.5 -6 × 2 -3.7 mm when flattened. Column curved, wings small and rounded, 3 -4 × 1.2 mm. Anther with a small triangular beak, 0.6 × 0.7 mm. Pollinia ovate, c. 0.2 mm diam. RECOGNITION. The species is distinct by the compact plant and inflorescence with the rachis longer than the peduncle, the semi-erect petals and the kneeshaped lip without a callus. Schlechter pointed out in his description that this species is related to Liparis capensis Lindl. (Lindley 1840: 314) known from the Western and Eastern Cape of South Africa. Although related, plants of L. capensis are generally shorter, normally have two narrower leaves, flowers about ¼ smaller, a slightly different lip (elliptic vs obovate) with a small conical callus at its base vs ecallose. There are similarities in plant habit to L. bathiei but the flowers are almost double the size, the lip is obovate vs ovallozenge shaped and the anther is triangular vs rounded. In plant habit it is similar to L. bicornis but the flowers are very different with the lip obovate vs ovate-obcuneate and without callus at the base vs distinctly two-horned. It is closest to L. lutea and they could well be the same species but the pseudobulbs are longer, the leaves bigger, the inflorescence and rachis more compact and the flower segments a little larger. Due to the limited nature of the type material of L. lutea it is impossible to effectively compare the two species. DISTRIBUTION. Endemic to Madagascar: Fianarantsoa, in the Andringitra massif and Toliara province in the Andohahela massif, at a similar altitude but several hundred km South from the type locality (Map 19).  Category EN: the extent of occurrence (EOO) of Liparis densa cannot be estimated since it is only known from two subpopulations whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). Liparis densa is only known from two subpopulations representing two locations (sensu IUCN), this rare species has thus been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. January. ETYMOLOGY. Presumably refers to the compact habit of plant and inflorescence. NOTES. The species was first collected by Perrier in the Andringitra Massif in SE Madagascar in 1922. It was described by Schlechter in Germany two years later and the type specimens were returned to Paris. There are two herbarium sheets of Perrier 14358 in P marked as 'type': the sheet with barcode P00094972 is the most complete with characteristic clumps of plants and is therefore chosen here as the lectotype, with sheet P00094972 being the isolectotype.
Humbert 13559 shows some differences from the type, as noted by Perrier: its pseudobulb and leafshape are similar, but its lip is a little larger and has a different shape but this is difficult to confirm in the dried specimens. ILLUSTRATIONS. Figs 19, 20. Schlechter (1932: T.52

RECOGNITION.
Liparis dryadum is one of the smallest species of the genus. The inflorescence is somewhat longer than the lanceolate leaves, of which it has two to three leaves with wavy margins and a short petiole.
The lip disk has a short, bilobed callus at the base. The broadly inverse-trapezoid lip with rounded corners is also characteristic.
Liparis dryadum is similar to L. xanthina but the leaves are shorter and narrower, the differences in flower characteristics are difficult to ascertain with only Deans Cowan's rudimentary watercolour as evidence. It is somewhat different from L. cladophylax which has wider leaves that are not corrugate at the margins and an narrower lip with an 'M' shaped callus vs shortly bilobed. Schlechter mentioned a similarity with L. epiphytica Schltr. (Schlechter 1905b) from West Africa, but it is easily distinguished by being more than half the size both in plant and flower and a bilobed lip callus vs rectangular.
Perrier seems to have made another collection in the same locality one month after collecting the type specimen and gave it the same collection number. The later collection is a topotype of the species (P00094978). It is also possible that an error was made on the specimen labels. The species was re-classified as Platystyliparis dryadum (Schltr.) Marg. by Margonska in Richardiana (2009: 94) but there is no evidence or necessity for this.  HABITAT. Epiphyte in lichen forest, or dense humid evergreen forest, moss and lichen-covered trees. Altitude: 660 -1600 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis dryadum is estimated to be 151,497 km 2 (far exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 20 km 2 (which falls within the limits for Endangered status under the criterion B2). With only four known subpopulations representing four locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. January to March. ETYMOLOGY. Referring to the Dryads, tree dwelling mythological nymphs. ILLUSTRATIONS. Figs 21, 22;Perrier (1939: 283); Cribb & Hermans (2009: 151). Very small to small, terrestrial or epiphytic plant, squat, 6 -11 cm, rhizome very short, older pseudobulb leafless and porrect to horizontal, the flowering one erect, roots glabrous, thin and wiry. Pseudobulbs 1 -5 cm, 8 -19 mm diam., covered by membranous leaf sheaths, carrying 2 -4 erectly-spreading leaves but generally with 2 prominent leaves. Leaves lanceolate or ovate, acute, 5.6 -6.5 × 1.2 -3.9 cm, membranous, normally strongly 7-veined, pale green. Inflorescence erect up to 22 cm long but usually shorter, thin, laxly 2 -15-flowered in the upper half. Peduncle carrying 2 -4 ovate-lanceolate, acuminate peduncle sheaths, slightly auriculate at the base, 3.  crenate at the margins, base with two small to very small rounded calli at the base forming a saddle-shape, 5 -6.3 × 4.9 × 5.5 mm. Column strongly arched, narrowed in the middle, with the wings obtuse-triangular, 1.8 -3.9 × 0.8 -1.2 mm. Anther rounded with a small acute apicule at the anterior margin, c. 1.1 × 0.9 mm. Pollinia ovoid, c. 0.4 × 0.3 mm. RECOGNITION. Liparis flavescens is a small and squat plant, with the pseudobulb covered by sheaths and leaf base, the previous year's pseudobulb being leafless and porrect. It bears three to four lanceolate to ovate, strongly veined leaves with a short petiole. Somewhat cordate, sterile bracts, auriculate at the base, are found along the lower half of spike. The medium to small flowers have a rounded lip, slightly crenate at front, with two small calli at base in a saddle shape and a short, sharply beaked anther.
Most, if not all records from Madagascar of Liparis flavescens are based on misidentifications. Liparis andringitrana seems to be its closest relative in Madagascar but generally its pseudobulbous stem is shorter and thicker, the leaf petiole longer, the flowers a little smaller and the callus more distinctly bilobed.
It is also similar to L. nephrocardia but it has a much longer inflorescence and rounded column wings (vs triangular in L. nephrocardia). It is close to L. nectarina in habit and lip shape but the leaves of L. flavescens are membranous not leathery, strongly many-veined (vs 3veined), the lip margin is entire (vs clearly emarginate and more deeply undulate), and the lateral sepals overlap (vs spreading). It shares the habit, inflorescence and flower habit of L. bowkeri from Southern Africa but L. flavescens is considerably larger, with more and larger flowers (Tables 2 & 3). DISTRIBUTION. Its main distribution is in Réunion and Mauritius but there is one record from the Seychelles: Moore in Baker (1877: 343) listed 'Horne!'. A specimen from Mahé at K, possibly Horne 602 is this species and is noted to be common in moist shady forest of Mahé, Praslin and Silhouette. The species is also listed by : 223), Summerhayes (1931 cited several specimens and recorded the species as common (Map 21). A record from the Comoros (Humblot 1509 at BM & P) is clearly not this species but Liparis sambiranoensis. Collections from Madagascar are generally of other species or difficult to ascertain because of the quality of the herbarium , this species has been preliminarily assessed as VU using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence slash-and-burn farming, grazing and anthropogenic fires, resulting in habitat reduction and habitat quality reduction. Previously assessed and published in the IUCN Red data listing for Réunion as LC (Picot 2013: 17) FLOWERING TIME. February to August. ETYMOLOGY. Refers to the yellowish colour of the flowers. NOTES Leptorchis. Ames (1908: 127) made the combination of Malaxis flavescens for an Asian species in error.
Du Petit-Thouars, in his description, indicated the width of the leaves at ¼ pouce wide (c. 6 mm), this is different from what is shown in his drawing. Perrier, in his revision of Liparis (1936: 256), was not certain of the existence of the species in Madagascar. He mentioned that a Lyall specimen quoted by Ridley (1885: 460) seems to be another species (this is L. imerinensis at K with a drawing in K-Lindl. ). Pervillé 710 (P), identified as L. flavescens, is in part L. foliosa: it seems to be a mixed sheet of different species, with duplicates at K. A collection by Neraud (G) from Réunion has a contemporary label of Malaxis flexuosa, but this is likely to be a transcription error. ILLUSTRATIONS the upper 1 3 of the inflorescence, becoming more lax as the flowers open. Floral bracts lanceolate, attenuate, more or less auriculate at the base, becoming smaller towards the apex of the rachis, 4.1 -9.8 × 0.8 -1.6 mm. Flowers small, erectly spreading 8 -10 × 6 -8 mm, characteristically the dorsal sepal strongly reflexed, the lip blade and lateral sepals in an almost vertical plane, petals and sepals pale yellowish green, lip darker green with a yellow-green disk becoming orange with age, column white, anther green. Pedicel and ovary erect, somewhat ridged, 9.5 -14 × 0.8 -1.8 mm. Dorsal sepal erect to strongly recurved, margins incurved ligulate, apex obtuse, 7.8 -10 × 1.5 -2.1 mm. Lateral sepals lanceolate with the apex narrowed and more strongly incurved, strongly down-curved and parallel with the blade of the lip, 6.1 -8.1 × 2.3 -3.9 mm. Petals recurved, margins incurved, linear, 6.3 -8.7 × 0.3 -0.6 mm. Lip ovate, strongly down-curved just below the base, base barely winged with two rounded, thickened ridges especially obvious in living plants, blade margins rounded more or less irregularly dentate, apex with a distinct protruding lobule, disk thickened, 5.8 -6.7 × 4.8 -5.5 mm. Column linear clavate, curved, the front margins with small angular wings towards the apex, 3.9 -6.4 × 0.8 -1.9 mm. Anther obovoid with a distinct rounded lobe at the anterior margin and sometimes with a miniscule sharp beak, c. 1 × 1 mm. Pollinia 2 fused pairs, narrowly ovoid, bipartite, c. 0.6 × 0.4 mm. RECOGNITION. A small to medium-sized plant, with distinctly rounded pseudobulbs, persistent, ligulatelanceolate leaves, a laxly flowered inflorescence and small flowers with reflexed segments, lateral sepals in a vertical plane with the lip, lip strongly decurved from the base, ovate with rounded ridges at the base, anterior margin lobed, disk thickened, darker green becoming orange, column with indistinct wings, anther roundly lobed or minutely beaked at the front margin.
Liparis foliosa is similar to the Australian L. reflexa but the lip is larger in L. reflexa, oblong (vs ovate), the anterior margin which is emarginate with a small central lobe in L. reflexa (vs a larger single lobe in L. foliosa). In Liparis reflexa the lip is thickened at the base into a single callus, whereas L. foliosa has thickened callus-like ridges, the orange colouring is linear on the former and covers the entire disc of the lip in the latter. It is similar to the Madagascan L. longipetala but the pseudobulbs in L. foliosa are larger, the leaves thicker, not longly acuminate, the lip callus smaller and the anther rounded and not beaked as in L. longipetala. DISTRIBUTION Hort. Barclay, 1824 (holotype K-Lindl.); Hort. Bot. Gard. Glasgow (paratype K-Lindl.). RÉUNION  the same sheet there is a preparatory pencil drawing for the illustration in the Lindley Herbarium at K and together with the plant material there is no doubt that this is the original material used by Lindley. The same year another illustration appeared in The Botanical Cabinet (Loddiges 1825(Loddiges : t.1097), here it is mentioned that the plant '…was lately introduced from Mauritius, by our worthy friend, Mr. Barclay, of Bury Hill, who kindly imparted it to us. It flowered in November, and is a curious and interesting plant, although not very showy…'. The text is presumably by George Loddiges, no author is mentioned of L. foliosa but one must assume it refers to Lindley's newly published name. Robert Barclay Esq. of Bury Hill (near Dorking in Surrey) (1751 -1830), a wealthy collector, grower and student of plants, was well connected in the botanical world and had a number of overseas contacts who supplied him with plant novelties. One of these contacts was Dr Charles Telfair (1778 -1833) a naval surgeon resident in Mauritius but also known to have collected plants in Madagascar, Réunion and the Seychelles (Dorr 1997: 468). W. J. Hooker refers to this in Barclay's obituary and also reproduced his last letter to Barclay (Hooker 1830: 122). Telfair also had a great interest in botany, working with Wenceslas Bojer, corresponding with several botanists in Europe and introducing a number of plants from the region. Loddiges cultivated several of them, presumably obtained via Barclay. The Horticultural Register of 1831 (Paxton & Harrison 1831: 190) recorded the species flowering at the Loddiges & Sons Nursery in Hackney. Another plant from the Barclay collection is illustrated and described in Curtis' Botanical Magazine in 1827 (t.2709) (Fig. 25) by W. J. Hooker who noted that it is '…drawn from a plant presented by Mr. Barclay, to the Glasgow Botanic Garden, which flowered in the month of Oct. 1826. It is a native of the Mauritius, and there, in all probability, grows upon trees. The figure given in Botanical Register being drawn from a young specimen, the bulbs were not formed, and hence, probably, Mr Lindley was induced to suppose the plant was terrestrial…'. There is a specimen that flowered at Kew, annotated by Robert Rolfe that may be the type of the illustration in Curtis' Botanical Magazine t.2709, there is another specimen at K of the plant that originally flowered at Glasgow Botanic Garden. Another plant of unknown origin, flowered in the Botanical Garden in Paris in the 1830s (P02118440) and was identified by Achille Richard as this species.
Considering the origin of the Liparis described by Lindley, there is no evidence that it actually came from Mauritius. Barclay may have noted the origin as having been sent by Telfair from Mauritius but it is possible that it could have been collected on Réunion. In 1828, Achille Richard included Lindley's species in his Monograph of the orchids of Mauritius and Réunion; he likened it to L. flavescens differing in its longer, narrower leaf and more flowers.
Five years after his original description, Lindley, in his Genera and Species of Orchidaceous Plants (1830a: 29) listed 'Liparis foliosa Lindley' but also has 'Liparis? reflexa Lindl. l.c.' and 'Cymbidium reflexum R. Br. Prodr.331' and had the species coming from Mauritius, New Holland and Port Jackson, the last two referring to an Australian origin. In his observations Lindley wrote that the plants from Mauritius were found to have an entire clinandrium and the Australasian one to have a dentate one but otherwise not different. This re-classification is explained further by Lindley in The Transactions of the Horticultural Society of London (1830b: 68), where he gave the same nomenclature arrangement and referred to a plant collected in New Holland by Captain McArthur in 1825 and described by Brown as Cymbidium reflexum. He goes on to say that 'Upon comparing it with the Ile de France Liparis foliosa, described by me in the Botanical Register, fol. 882, I see no reason to doubt its being the same species; that plant which flowered in Mr Barclay's stove, and which I there described, was destitute of bulbs, but the plant of the Garden was furnished with large ovate oblong bulbs, each of which was terminated by three ligulate acuminate channelled leaves indistinctly five-nerved. The scape was about the same length as the leaves of that shoot which produced it, but much shorter than the full grown leaves of the old bulb. The differences that seem to exist I am disposed to believe depend upon the age of the plant, which, when old, would have bulbous stems, of which it would be destitute when young …' De Candolle in his Géographie Botanique Raisonnée (De Candolle 1855: 1036) used the variability and distribution given by Lindley for L. reflexa, to illustrate the geographical variance in species. Lindley's decision of putting the Mauritius plant into synonymy with Brown's Australian L. (Cymbidum) reflexa was followed by subsequent authors, including Ridley's monograph of the genus (1887: 287).
Moore in Baker (1877: 343) in the Flora of Mauritius and the Seychelles noted that 'Liparis foliosa Lindl. given as Mauritius on the authority of Barclay, is in all probability, not a native of this region. Du Petit-Thouars makes no mention of it, and it was not seen by either Richard or Bojer. On the other hand, carefully dissecting it will, I believe, show that it is identical with the Australian L. reflexa Examining the literature and all the material available it is clear that Liparis foliosa, as described by Lindley in 1825 is a valid species and is different from L. reflexa (R.Br.) Lindl. from Australia, SE Queensland to E New South Wales.
The possible confusion of the origin of Barclay's plants is questionable as the provenance from Telfair is solid and there are no records of Barclay importing plants from Australia (as explained above and also see (Hooker 1830: 122)); the only question is over its origin from Mauritius or Réunion. The cultivated plants from Kew and Glasgow have the same origin, as have the illustrations in the Botanical Register, Botanical Cabinet and Curtis' Botanical Magazine.
There are undoubted similarities between Liparis foliosa and L. reflexa from Australia; they both have bulbous pseudobulbs, linear-lanceolate leaves, greenish-yellow flowers with orange marking on the lip and reflexed sepals and petals but so do several other Liparis from around the world. It is not surprising that Lindley thought them to be the same species, at a time when only limited material was available. There are a number of differences between the two species (Table 9): the plant habit is somewhat different with the pseudobulbs more elongate in L. reflexa, the leaves and inflorescence longer. The principal differences are in the oblong lip (vs ovate) which is larger in L. reflexa, the anterior margin which is emarginate with a small central lobe in L. reflexa (vs a larger single lobe in L. foliosa), a single thickened callus in L. reflexa (vs thickened callus-like ridges), its orange coloured stripe (vs entirely orange disc). The column of L. foliosa is more acutely winged.
The A similar specimen has been recorded from Madagascar (Lavango Surveillant 6031 in P) but as the specimen has seed capsules only it is not possible to ascertain its identity. ILLUSTRATIONS  Medium to large slender plant, erect terrestrial,  RECOGNITION. It is likely that this species has been confused with other vegetatively similar species with a long stem but the medium-sized flowers with a distinct rounded lip with wavy edges, strong side veins and the dentate callus are characteristic.
In plant habit the species is similar to Liparis chantaliae (described above) but the leaves are much thicker in texture and elliptic vs broadly oval, the lip suborbicular vs broadly oval-lanceolate and the callus tridentate vs bi-lamellar, the flowers are also much smaller. It equally has similarities with L. longicaulis with its elongate stem but differs in that the stem is very thin and has two to rarely three leaves with short wavy margins. The flowers are distinct by the wide rounded lip with at the base an erect three pointed wedge-shaped callus. The species is also similar to L. caulescens from Réunion, but the leaves are more cordate, there are fewer flowers a quarter smaller and the lip is suborbicular vs broadly obovate with a tridentate vs bidentate callus. DISTRIBUTION  status under the criterion B2). With only four known subpopulations representing four locations (sensu IUCN), L. gracilipes has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. The habitat near Anosibe an'ala has been destroyed due to slash-and-burn. FLOWERING TIME. January to May. ETYMOLOGY. Referring to the long slender pseudobulbs. NOTES. The species was described by Schlechter in Germany from herbarium material sent to him by Perrier and this was then returned to the Paris herbarium, as indicated on the herbarium label. In his description, Schlechter cited Perrier 15747 as the type, there are two sheets of this number and P00094979 is designated here as the lectotype and P00094980 is an isolectotype; the first specimen is more representative of the species and has more open flowers. Until recently known from the type only. ILLUSTRATIONS. Perrier (1939: 255).

mm.
Anther triangular acute at the front, up to 1.3 × 1.1 mm. Pollinia oval c. 0.5 × 0.6 mm. RECOGNITION. Its single erect flowering growth reaches up to 25 cm, with the leafy pseudobulb up to 11 cm long, the other leafless and leaning to almost horizontal. Lip oval-rhomboid, with a bidentate callus, acute at the apex, with five veins and anther with a short beak at the front.
Similar to Liparis andringitrana in some respects but it is a somewhat larger plant with longer pseudobulbs, the older pseudobulb being almost divaricate. Its lip has an acute tip whilst that of L. andringitrana is rounder and indented, the callus consists of two horns (vs a rounded lobe). DISTRIBUTION HABITAT. Epiphyte, terrestrial or lithophyte amongst moss, moss forest, evergreen forest. Altitude: 1200 -1800 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis imerinensis is estimated to be 53,247 km 2 (far exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 40 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis imerinensis is known from nine subpopulations representing nine locations (sensu IUCN), thus been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. The habitat in the Manjakandriana region has been destroyed and is now a Eucalyptus plantation. FLOWERING TIME. December to May. ETYMOLOGY. Referring to Imerina; an old kingdom and geographical area of the highlands of Madagascar. VERNACULAR NAMES. 'Famany' (d'Alleizette 346 in P). NOTES. Schlechter (1924) described Liparis imerinensis from a collection made by Perrier in 1920 in Manjakandriana to the E of Antananarivo. He mentioned that he had only little material available for his description. Perrier (1939: 294) added further information from other collections for his description in the Flore de Madagascar. Schlechter cited Perrier 12973 as the type; there are two sheets of this number in Paris and P00094997 is designated here as the lectotype and P00094998 is an isolectotype, the first specimen being more complete.
Liparis imerinensis seems to be a variable species both in plant and floral characteristics: the callus on the lip is obvious in some collections but hardly noticeable in others . Perrier 16869, 18431 and 19010 (all P) are all annotated by Perrier as 'var', perhaps confirming the variability of the species. ILLUSTRATIONS. Fig. 27;Schlechter (1932: t.53), the drawing does not show a callus on the lip; Cribb & Hermans (2009: 157); the photograph in Bosser & Lecoufle (2011: 408) shown as Liparis imerinensis has the characteristics of L. ochracea.
Small to large erect terrestrial plant, 8 -24 cm high, rhizome very short or absent, roots wiry, somewhat woolly. Pseudobulbs 2, stem-like, elongate, fleshy, the old one leafless and at a wide angle to the flowering one or lying flat, enveloped by 2 -3 brown sheaths, sometimes with a large leaf-like sheath, up to 12 cm long, c. 5 mm diam. Leaves 2 -3, erectly spreading, thin, oval to obliquely elliptic, shortly acuminate, narrowed at the base into a short pleated-veined, petiole, 4 -13 × 1.5 -6.5 cm. Inflorescence slender, erect, laxly 4 -9-flowered. Peduncle up to 9 cm, with a few sterile oval lanceolate bracts, av. 7 × 3 mm. Rachis c. 5 cm long. Floral bracts erectly spreading, lanceolate, around half the length of the ovary, 4.5 -7 × 2 -3 mm. Flowers medium size, pale greenish-yellow becoming brown with age, 12 -15 × 9.0 -12 mm. Pedicel and ovary rounded at the base becoming angular towards the pedicel, up to 15 × 1.3 mm. Dorsal sepal recurved, lanceolate, subacute, 7.1 -11 × 1.5 -2.1 mm. Lateral sepals porrect, semi-oblong, somewhat obtuse, 7.1 -9 × 3.5 -5.1 mm. Petals decurved, narrowly linear, subacute, 7.1 -9.0 × 0.3 -0.6 mm. Lip transversely elliptic, with proportionally small wings at the narrow base, the blade very rounded at the sides, the front margins almost straight ending in a small triangular cusp, with two small but distinct thickened calli at the base thinning out into three strong central veins, 5.1 -6.3 × 6.9 -10.1 mm. Column almost straight, somewhat broadened towards the base, with two small wings at the apex, 3.0   Map 25. Distribution of Liparis jumelleana. RECOGNITION. Usually a squat plant with 2 -3 thin leaves on the flowering growth, the previous growth being leafless and lying at an angle or flat on the ground. It has short, stem-like pseudobulbs, mediumsized flowers with a typically broad lip with a short triangular apex, small wings at the base and two small but distinct nipple-like calli at the base thinning out into the three central veins, and an anther with a sharp beak.
The only other species with a very broad lip is Liparis laurentii described below but L. laurentii has a single perennial leaf on a slender pseudobulb (vs 2 to 3 deciduous leaves of a different shape and size), the lip calli are shorter and blunter in L. jumelleana. In habit it somewhat resembles L. imerinensis but the lip shape and callus are very different with the lip transversally elliptic vs oval and the callus small and rounded vs two-horned. DISTRIBUTION HABITAT. Epiphyte, terrestrial or lithophyte in humid evergreen forest at mid-altitude. On gneiss rock. Altitude: 300 -2200 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis jumelleana is estimated to be 4,675 km 2 (within the limits for Endangered status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 20 km 2 (which also falls within the limits for Endangered status under the criterion B2). With five known subpopulations representing five locations (sensu IUCN), L. jumelleana has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. October to May. ETYMOLOGY. Named for Henri Jumelle (1866 -1935), professor of botany and director of the Botanic Garden in Marseille (France). He did not work in Madagascar but collaborated extensively with Perrier. NOTES. Some of Perrier's earlier collections followed an alternative numbering system: here the type has the old number 133 bracketed by Perrier, with the new number 11366. As often, Schlechter assigned more than one herbarium sheet to the name; from these the specimen P00095450, corresponding most closely to the description, has been designated as lectotype, the other (P00095451) as the isolectotype.
Plant size is quite variable as is obvious from the type specimens but lip size and shape seem to be consistent throughout its range. ILLUSTRATIONS. Fig. 28;Schlechter (1932: t.54); Perrier (1939: 283). Small scrambling to ascending terrestrial plant of 2 -3 growths, up to 9 cm high, on a short rhizome, roots flexuose, somewhat woolly, c. 1 mm diam. Pseudobulbs slender, ascending, tubular, more or less thickened towards the base, 20 -45 × 5 -10 mm, dark green, entirely enveloped by 2 -3 long brownish papery attenuate sheaths, strongly ridged and with a few dark brown specks, the upper one caudate, up to 23 × 7 mm, with a single leaf at the apex. Leaf somewhat fleshy, persistent on the older pseudobulbs, broadly ovate, acute, rounded to subcordate at the base, with a 1 -2 cm sheathing petiole, blade leathery and rigid, pale green, with several darker veins, 3.1 -7.5 × 2.5 -4.5 cm. Inflorescence erect from the apex of the pseudobulb, up to 9 cm long, carrying up to 5 flowers but generally fewer. Peduncle partly enveloped by the leaf petiole and with 1 -2 bract-like sheaths, 5.1 -9 × 3.2 -4.6 mm. Rachis loosely racemose, the older flowers fading before the new ones open. Floral bracts lanceolate, somewhat cordate at the base, 3.5 -6 × 0.9 -3.1 mm. Flowers medium sized, c. 1.8 × 1.5 cm, facing upwards, the ovary and floral bracts pale whitishgreen, the petals and sepal brownish green, the lip pale green with the disk darker viridian, anther darker green. Pedicel and ovary slender, slightly winged, sometimes somewhat dentate, 10 -16 × 1.1 -2.1 mm. Dorsal sepal erect, arching, the margins recurved, ligulate, acute, 7.2 -10.3 × 1.1 -1.8 mm. Lateral sepals falcate-oblong, obtuse, parallel with the lip, margins a little recurved, 5.1 -9.8 × 3.2 -4.8 mm. Petals spreading, curled upwards towards the apex, margins recurved, linear, 7.2 -9 × 0.3 -0.5 mm. Lip transversely elliptic to reniform, with small basal wings, the margin entire but the anterior tip emarginated with a small central lobule, base with two distinct hornlike calli, 4.1 -6.5 × 6.1 × 7.8 mm. Column slightly arched towards the apex, wings long, membranous, rounded, 2.5 -3.9 × 1 -1.3 mm. Anther ovate with a distinct sharp beak upturned at the apex, sometimes a little verrucose on the exterior, remaining attached by a thin filament at the base after pollination, c. 2.2 × 0.8 mm. Pollinia ovate c. 0.4 × 0.3 mm. Fruit: young capsule ovoid, 7 × 4 mm, with six longitudinal bands including three more protruding, mature capsule and seeds unknown. RECOGNITION. Liparis laurentii is a small single-leaved plant with a slender pseudobulb, broadly ovate leaf, few-flowered inflorescence, and medium-sized flowers, with a transversely elliptic to reniform lip, the anterior tip emarginate with a small central lobule and a base with two distinct horn-like calli.

Liparis laurentii
It is related to Liparis clareae but smaller, the leaf has a more distinct central vein and the base is rounded (vs cordate), has fewer flowers (1 -5 vs 6 -12), a much wider, transversely elliptic to reniform lip (vs ovate), with entire margins (vs serrate), an emarginate apex (vs entire), horn-like basal calli (vs slight swellings) and elongate column wings (vs short and obtuse). It has similarities with L. warpurii but the leaf of L. laurentii is broadly ovate (vs lanceolate), the flowers smaller and the lip not serrate at the margins and with horn-like calli (vs a single rounded callus). It is similar in size to L. magnifica described below but the lip is transversally elliptic (vs obovate), the anther beak is much longer and the lip callus is separate (vs joined). It differs from L. superclareae described below, by its diminutive size, and shape of the lip, callus and anther (Table 8). The only other species with a very broad lip is L. jumelleana but L. laurentii has a single perennial leaf on a slender pseudobulb (vs two to three deciduous leaves of a different shape and size on a ridged pseudobulb), and the lip calli are shorter and blunter in L. jumelleana. DISTRIBUTION HABITAT. Terrestrial amongst leaves and humus in forest from 410 -1293 m altitude. In the Loky-Manambato National Park (Daraina), a total of about 40 individuals were observed in one locality during the flora and vegetation survey of the region. The species was there growing on the ground, along a crest, close to the top of a granitic mountain reaching 607 m elevation with a non-climatic raw mineral soil (Riquier 1968;Jourde et al. 1974). It grows in a mesophilous forest surrounded by a matrix of dry deciduous forest. The vegetation of the Daraina locality reaches up to 10 -15 m high with emergent trees up to 18 (-21) m, with a quite dense shrub stratum at 0.5 -1 m high, with shrub strata increasing in density between 4 -8 m high and with a very sparse stratum at 8 -10 m high (Gautier et al. 2006;Nusbaumer 2011 Category EN: The extent of occurrence (EOO) of Liparis laurentii cannot be estimated since it is only known from two subpopulations, one included in the Loky-Manambato Category V Protected Area (Daraina) and the second one at 100 m exterior of the limit of the provisionally protected area of the Corridor Marojejy-Anjanaharibe-Sud-Tsaratanana (COMATSA), whereas its minimal area of occupancy (AOO) is estimated to be 8 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). With only two known subpopulations representing two locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. In the Daraina region, this species was collected rather early during the floristic survey and we expected several subpopulations to be observed later, because its habitat, the mesophilous forest, is quite frequent in Loky-Manambato. Nevertheless, it turned out that no additional fertile or sterile material of this species was encountered among the more than 54,000 plant species occurrences recorded including about 5,000 herbarium specimen collections (Nusbaumer 2011  ETYMOLOGY. Named for Dr Laurent Gautier of the Conservatoire et Jardin botaniques de Genève, one of the co-finders of the new species, the upturned petals also have a striking resemblance to his beautifully groomed moustache. NOTES. The new species is known from two localities in North-Eastern Madagascar. Specimens in herbaria had either been unidentified or named as Liparis clareae with which it shares a number of features but is very different, as discussed above. The plant from the Doany area (Ravelonarivo et al. 1606) comes from a much higher elevation than those collected in the Bobankora forest near Daraina (Gautier et al. 4204), they also seem to consist of more mature and therefore larger plants and flowers; all other characteristics and proportions are otherwise identical. ILLUSTRATIONS Medium sized, slender, erect terrestrial or epiphytic plant, up to 18 cm high, rhizome very short, roots wiry, more or less hairy. Pseudobulbs stem-like, thin, erect, fleshy, enveloped by thin grey striate sheaths overlapping each other, with two leaves at the apex (rarely 3), older growths retaining their leaves, up to 13 cm, c. 5 mm diam. Leaves sub-opposite, erectly spreading to sub-erect, oval or elliptic-lanceolate, acute, base rounded or wedge-shaped, sessile or on a short petiole, fairly leathery and glossy, pale green becoming darker with age, 3.5 -7.3 × 1.5 -3.0 cm. Inflorescence erect, up to 15 cm long and almost as long as the pseudobulbs, sub-laxly 5 -15-flowered. Peduncle almost as long as the raceme, slightly ridged, with two to three sheaths c. 5 × 3 mm. Rachis successive flowering, up to 9 cm long. Floral bracts erectly spreading, acuminate, about 1 3 -½ the length of the pedicellate ovary, c. 4 × 2 mm. Flowers small to medium in size 8 -11 × 6.5 -7 mm, greenishyellow becoming more yellow with age, the disk of the lip darker viridian green, lip generally appears darker in herbarium material. Pedicel and ovary glabrous, slender, a little ridged, 8 -12 × 0.7 -3 mm. Dorsal sepal reflexed (often remaining in herbarium material), narrowly lanceolate almost tubular, somewhat obtuse, base subcordate to auriculate, 5.5 -8.0 × 1.1 -1.5 mm. Lateral sepals folded beneath the lip, semi- oval, obtuse, 5.0 -6.0 × 2.5 -3.9 mm. Petals reflexed, linear almost tubular, subacute, slightly narrowing towards the base, 5.5 -8.3 × 0.5 -0.8 mm. Lip slightly curved, shortly auriculate at base, expanded into an obcordate blade sometimes covered by small viscous droplets in living plants, with two simple central veins and two branched lateral ones, the margins curved, the front margin a little denticulate-undulate, subtruncate or emarginated with a bicornate or 2horned callus at the base which can be up 1.4 mm long, 4.5 -7.1 × 5.0 -6.9 mm. Column slightly curved, up to 5 × 1.5 mm, somewhat angular wings. Anther with a short, obtuse beak, c. 1.0 × 1.1 mm. Pollinia 2 pairs, oval c. 0.5 × 0.6 mm. RECOGNITION. Distinguished by its long stem with two somewhat leathery oval-lanceolate leaves at the top. Its lip is entire, slightly indented on the anterior margin and rounded with a distinct two-horned callus at the base. Its anther is slightly beaked. It is a very variable species both in shape of the lip and that of the two calli on the lip which can be more or less united or separate and up to 1.4 mm long, parallel or slanting. In herbarium specimens, the lip of the flowers often appear considerably darker and denser than the rest of the floral segments.
Both Liparis bernieri and L. danguyana are similar species but there are substantial differences ( Table 5). The leaf of L. danguyana is thinner than those of L. bernieri and L. listeroides which are more leathery with indistinct veins. The flowers of L. listeroides are generally smaller than the other two. The lip of L. danguyana has two small but distinct calli at the base, those of L. bernieri are less distinct and rounded, those of L. listeroides are long and tooth-like. The front margin of the lip of L. danguyana is apiculate, the others have a slightly d e n t i c u l a t e m a r g i n . T h e a n t h e r c a p o f L. danguyana is rounded at the front and suborbicular, that of L. listeroides is slightly beaked. Considering these differences and specific geographical distribution it is reasonable to recognise all three species. Vegetatively, there is some resemblance with L. stenophylla but the leaves are more lanceolate than elliptic. It shares the basal teeth of the lip with L. polycardia and L. imerinensis but the plant and flower are smaller, the leaves shorter and broader and the inflorescence shorter. DISTRIBUTION Mt Tsaratanana, forest, c. 1500m, April 1924; Centre, Tsinjoarivo, forest, c. 1400m, Feb. 1925, Perrier 16970bis (P); Mandraka, undergrowth, evergreen forest, Feb. 1959, Bosser 12625 (P); Mandraka, Feb. 1959, Bosser 12723 (TAN); Antananarivo, Mandraka, PK69, March 1984, Dorr et al. 2918; E of Ankaramy, Manongarivo, Antsatrotro, 1470-1570m, April 1992, Malcomber et al. 1498  HABITAT. Moss forest, evergreen forest in undergrowth, in shade, moss and lichen-covered trees. Altitude: 1040 -1600 m. CONSERVATION STATUS. Category LC: the extent of occurrence (EOO) of Liparis listeroides is estimated to be 87,271 km 2 (far exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 56 km 2 (which falls within the limits for Endangered status under the criterion B2). Because L. listeroides is known from 12 subpopulations representing 14 locations (sensu IUCN), this species has been preliminarily assessed as LC using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. More specifically the species has been affected by habitat loss due to land clearance by burning in the Manongarivo and Mandraka areas. Also affected by charcoal production in the Lakato area. FLOWERING TIME. January to April. ETYMOLOGY. The plant resembles a medium-sized Listera R.Br. in its habit (Schlechter 1924: 143). Listera, in turn, was named for Dr Martin Lister (c. 1638 -1712), English physician and naturalist. NOTES. Schlechter assigned more than one herbarium sheet to the type; the specimen (P00095453) corresponding most to the description and most complete was designated as the lectotype, the other (P00095454) being an isolectotype. ILLUSTRATIONS. Fig. 31;Schlechter (1932: t.54); Cribb & Hermans (2009: 151); Hervouet (2018: 410).
Flowers large, up to 24 × 18 mm, olive green to dark green becoming more yellow with age, the lip with a darker disk, column and ovary white, pollen yellow, flowers opening in succession, the upper ones sometimes cleistogamous or deformed. Pedicel and ovary somewhat ridged to winged, 11 -17 × 0.6 -1.8 mm.
Pollinia two pairs united at the tip by an obsolete viscidium, c. 0.7 × 1 mm. RECOGNITION. This is a large plant with long angular (sometimes not obvious in herbarium material) pseudobulbs, almost entirely covered by white sheaths, broadly ovate leaves, and large flowers with an obtuse lip, crenulate at the front, and a small rounded callus at the base. It bears some similarity to Liparis gracilipes but its leaves are much thinner and broadly oval vs elliptic, the lip ovate vs suborbicular, the callus indistinct vs tridentate. There is also some similarity to L. zaratananae but again the pseudobulbs are angular vs rounded, the leaves broadly ovate vs ovate-elliptic and the lip ovate vs oval. Historically it has been confused in the field and in herbaria with the newly described L. chantaliae but it is distinct because of its angular (vs rounded) pseudobulbs covered by papery sheaths, short rhizome, slightly smaller flowers, lip with an obtuse tip (vs attenuate), indistinct callus and the less acute beak of the anther (Table 7). DISTRIBUTION. Widespread and endemic to Madagascar: Antananarivo, Antsiranana, Fianarantsoa, Mahajanga and Toamasina provinces (Map 28). HABITAT. Evergreen forest, moss forest, riverine forest, degraded forest. Often amongst dense undergrowth in deep shade in humus-rich and moist soil. Altitude: 800 -1800 m. CONSERVATION STATUS. Category VU: the extent of occurrence (EOO) of Liparis longicaulis is estimated to be 95,273 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 104 km 2 (which falls within the limits for Endangered status under the criterion B2). With 11 subpopulations representing 17 known locations (sensu IUCN), this species has been preliminarily assessed as VU using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. December to April. ETYMOLOGY. Referring to the long distinct pseudobulb-stem. NOTES. Henry Ridley (1885: 461)

Ridley based the name on both collections. The Deans
Cowan material is duplicated in P but both sheets are less typical than that of Hildebrandt which has five collections in different European herbaria. The latter was therefore chosen as the lectotype for the species. Perrier in 1924 annotated a specimen from Mt Tsaratanana in N Madagascar (16072) as Liparis longicaulis var. crassilabia, mentioning differences from the Imerina (Highland) forms in the lip lacking visible veins and the anther having a very short rostrum. This varietal name was never published. The plant shows the typical characteristics of the species but the lip is a little broader and more acute at the apex.
Liparis longipetala Ridl. (Ridley 1885: 459;1886b: 275);Perrier (1936: 243); Hermans et al. (2007: 220) ;Cribb & Hermans (2009: 152 cence and flowers with a long rounded callus at the base of the lip, long and angular column wings, and an anther with a very acute beak. The lip often appears darker than the rest of the flower in herbarium material. It is similar to Liparis foliosa from Réunion but the pseudobulbs of L. longipetala are smaller, the leaves longly acuminate, the lip callus more pronounced and the anther beaked and not rounded as in L. foliosa. The flower is somewhat similar to those of L. bulbophylloides but the plant is caespitose vs longly rhizomatous, the leaves ligulate vs ovate, and the flowers almost half the size. DISTRIBUTION. Endemic to Madagascar: mainly in the Eastern forests of Antananarivo, Toamasina and Fianarantsoa province but also in the North in Antsiranana province. The Ankafana area referred to in the type is the old mission post near Fianarantsoa (Map 29) . Szelengowicz & Tamon (2013: 361), cited and illustrated this species from Réunion but its habit, leaf and shape of the lip are different; the plant illustrated belongs to a different species and is probably Liparis foliosa.  CONSERVATION STATUS. Category LC: the extent of occurrence (EOO) of Liparis longipetala is estimated to be 117,936 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 52 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis longipetala is known from seven subpopulations representing 12 locations (sensu IUCN), this species has been preliminarily assessed as LC using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. February to August. ETYMOLOGY. Referring to the petals which are said to be longer than in other species. Liparis anthericoides refers to the vegetative similarities with Anthericum (Liliaceae). NOTES Perrier 17103 was collected at 1300 m and has a good range of plants and also has small drawings by Perrier of floral detail, it is therefore designated as the lectotype. Perrier 17104 comes from the same locality but was collected at c. 1000 m and is a lectoparatype. He also cited Perrier 10981 in his description and this is also to be treated as a lectoparatype. There is no doubt that L. longipetala and L. anthericoides are the same which unfortunately means that Perrier's appropriate epithet, lacking precedence, becomes synonymous with Ridley's less appropriate epithet (Table 10). Plant size of this locally common species is variable but the shape of pseudobulb, leaves and inflorescence are consistent. The lip-shape with its characteristic narrowed base, oblong blade and long rounded ridge-like basal callus with a more or less bilobed end are identical, the sharply beaked anther and small triangular pointed column wings are also the same. It is unlikely that Perrier saw any of the type material of L. longipetala and he repeated some of Ridley's limited description in his own texts (Perrier 1936(Perrier : 243 & 1939. He mistakenly Liparis longipetala is a widespread species in the humid evergreen forest of Madagascar above 800 m and its flowering time from February to August is quite long. Ridley cited long petals as one of the characteristics of his L. longipetala but these are variable and not unusually long for the genus. ILLUSTRATIONS. Figs. 34, 35 and 36;Perrier (1939: 255); Cribb & Hermans (2009: 147); Bosser & Lecoufle (2011: 405); Guérin & Hervouet (2013: 320); (Hervouet 2018: 405).
Liparis lutea Ridl. (Ridley 1885: 458;1886b: 274);Perrier (1939: 279); Hermans et al. (2007: 220) ;Cribb & Hermans (2009: 152 RECOGNITION. A very small plant with a dense raceme, the flowers not spreading much and almost appressed to the rachis, relatively large floral bracts, and small, yellow flowers with lip with three central veins, lacking a callus, a slightly winged column and heart-shaped anther. The type material and the drawing by Deans Cowan (Cowan 1880: 22) provide limited evidence of the morphology of the species; it is similar to Liparis densa but the pseudobulbs are shorter, the leaves smaller, the inflorescence and rachis more elongate, carrying at least two lanceolate sheaths low on the peduncle (not present or smaller in L. densa) and the flower segments smaller. Due to the limited nature of the type material of L. lutea, it is difficult to compare the two species. It is also similar to L. bathiei but the floral bracts are smaller and narrower, the anther is beaked (vs cordiform) and the column has more pronounced wings. DI STRIBUTION. Endemic to Madagascar: the Fianarantsoa area only (Map 30). The type locality of Ankafana refers the highland area in-between Antsirabe and Ambalavao.  greenish brown to dark red, 12 -51 × 2 -7 mm, almost completely covered by 2 -3 papery, partly overlapping sheaths which are strongly acuminate at the apex, 11 -23 × 3 -5 mm, carrying a single semierect to divergent leaf at the apex. Leaf persistent, lanceolate, rounded to subcordate at the base, attenuate at the tip, pale to dark green, distinctly marked with silvery-white longitudinal interrupted streaks or continuous bands, 3 -5 × 1.3 -2 cm. Inflorescence erect, apically racemose, with 1 -5 flowers, 4 -6.5 cm long.
Peduncle somewhat angular to corrugate, about 2 =3 of the inflorescence, pale green, with 1 -2 lanceolate peduncle sheaths up to 7 × 2 mm. Rachis apical, extending as the flowers open. Floral bracts narrowly lanceolate, 3.6 -5.1 × 1.1 -1.7 mm. Flowers medium in size, erectly spreading, up to 14 × 9 mm, petals and sepals whitish-green to pale green becoming yellowishgreen when maturing, lip green with the disk, column white. Pedicel and ovary ridged, 7.5 -10.3 × 0.6 -0.9 mm. Dorsal sepal erect, recurved towards the apex, ensiform, margins recurved, 6.9 -8.2 × 1.2 -1.9 mm. Lateral sepals strongly recurved beneath the lip, elliptic, acute at the tip, 5.9 -6.5 × 2.5 -2.9 mm. Petals parallel or slightly divergent to the pedicellate ovary, linear, margins recurved, 7. RECOGNITION. This small new species is distinct by the single lanceolate leaf, marked with silvery white longitudinal bands, the short pseudobulbous stem, lanceolate floral bracts, the obovate lip with an undulate anterior margin and a distinct bidentate callus at the base. The species is distinct from all the other single-leaved Liparis of the region by the longitudinal silvery markings on the lanceolate leaf. It is closest to L. warpurii but differs by the lanceolate peduncle sheaths and floral bracts narrowly lanceolate vs cordate in L. warpurii, the leaf with an overall silvery upper surface vs banded, the flowers about half the size, the lip is more curved and obovate vs angular, the lip margin undulate vs dentate-serrate and the more sharply beaked anther. There are also similarities with L. clareae but the new species is much smaller in both plant and flower and the callus is bidentate vs indistinct. It is similar in size to L. laurentii described above but the lip is obovate vs transversally elliptic, the anther beak is much shorter and the lip callus is joined vs separate, it differs from L. superclareae described below by the much smaller plant and flower, the shorter anther and the lip callus bidentate vs obscure (Table 8). DISTRIBUTION species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. April and May. ETYMOLOGY. The epithet refers, firstly, to Madame Mag Izouard of Vakôna Lodge in Madagascar, who brought this species to our attention and, secondly, to the magnificence of the markings on the foliage. NOTES. For some time the new species was only known from photographs of its leaves which had been found in forest around Vakôna Lodge near Andasibe in Eastern Madagascar by its owner Mme Izouard. On recent flowering it was possible to match the species to several herbarium collections from the same area.
Since then the new species has also been found in the Ranomafana area in the SE of the island. Liparis cardiophylla var. angustifolia, mentioned by Perrier (1936: 244) and only known from the type specimen, also has whitish leaves with darker veins. It was initially thought that it could be the same as L. magnifica but its flowers are very different and correspond well with the characteristics of L. warpurii. It is discussed under that species. ILLUSTRATIONS Very small to small epiphytic or terrestrial plant 4.5 -10 cm high, on a very short rhizome, roots wiry, slightly villous, c. 1 mm diam. Pseudobulbs ovoid to elliptic 15 -20 × 7 -10 mm, covered by 3 -4 whitish brownish amplectant, scarious sheaths, with 2 -3 leaves, the third smaller and sheath-like number of leaves, the older pseudobulbs without leaves. Leaves erectly spreading, elliptic-lanceolate, in living plants corrugate into three distinct veins, in herbarium material several-veined, overall 3.2 -7.5 × 1 -1.9 cm, acuminate, with a short 5 -12 mm petiole, somewhat leathery, pale green. Inflorescence erect, relatively thick for the genus, corrugate up to 12 cm long but generally shorter, c. 2 mm in diam., with 5 -7 flowers. Peduncle about 2 =3 of the inflorescence, with 1 -2 elongate peduncle sheaths 8 -15 × 3 -5 mm. Rachis densely racemose becoming more spreading as the flowers open. Floral bracts lanceolate, acuminate, 6 -8.5 × 1.2 - 1.8. Flowers parallel with the rachis, small to medium in size, overall c. 11 × 7 mm, segments pale green becoming yellow and then orange with age, anther paler, the lip darker green with the central ridge and callus almost olive-green. Pedicel and ovary elongate, twisted, sharply ridged, 7.1 -8 × 1.3 -1.8 mm. Dorsal sepal erect, narrowly lanceolate, acuminate, the margins infolded, 7.8 -8.1 × 1.2 -1.8 mm. Lateral sepals ovate, obtuse, folded beneath the lip, spreading and only overlapping near the base 5.6 -5.8 × 2.6 -3.8 mm. Petals descending, more or less curved, linear, margins infolded, 8.2 -9 × 0.6 -0.9 mm. Lip distinctly auriculate at base, then very strongly curved and expanded into an obcordate blade, the margins curved, the anterior margin undulate, broadly emarginated, a bilobed rounded callus at the base becoming a longitudinal swollen ridge on the disc, covered by small sticky droplets in living plants, overall 5.5 -6.1 × 5.3 -6.4 mm. Column curved and roundly bilobed at the apex, 3.8 -4 × 1.4 -1.5 mm. Anther oval with a distinct acute beak at the anterior margin, 0.9 -1 × 0.9 -1 mm. Pollinia (2) ovoid c. 0.5 × 0.4 mm. Seed capsule obovate, erect, c. 6 × 5 mm. RECOGNITION. Liparis nectarina is a small plant with an ovoid pseudobulb covered by sheaths, two ellipticlanceolate corrugate leathery leaves with three distinct recessed veins, an inflorescence with one or two elongate sheaths, few-flowered towards the tip, and  small to medium-sized flowers, with a lip strongly auriculate at the base, with an emarginate apex, undulate, with a rounded bilobed callus and swollen longitudinal disk, nectar-like droplets (in living plants) on the surface, and a sharply beaked anther.
Liparis nectarina is closest to L. flavescens in habit and lip shape but the leaves of L. nectarina are leathery (vs membranous), strongly three-veined (vs manyveined), the lip margin is clearly emarginate (vs entire) and more deeply undulate. It is also similar to L. scaposa but the flowers of L. nectarina are at least a third larger, it has two leaves (vs one) and the anther is sharply beaked (vs roundly lobular). The lateral sepals overlap in both of the above species whilst they are divergent in L. nectarina, the distinct viscous droplets on the lip also seem to be particular to this species. DISTRIBUTION. Endemic to Réunion. Frappier refers to 'Hab. Saint-Pierre (Tampon)'. It has now also been found in nearby localities at the Entre Deux and Takamaka in the S Central mountains of the Island. Szelengowicz & Tamon (2013: 363) also report it from the more Westerly Cirque de Mafate (Map 32). be 164 km 2 (which falls within the limits for Endangered status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 16 km 2 (which also falls within the limits for Endangered status under the criterion B2). Liparis nectarina is known from three subpopulations representing two locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by grazing and anthropogenic fires, resulting in habitat reduction and habitat quality reduction. Previously assessed and published in the IUCN Red data listing for Réunion as category DD (Picot 2013: 25). FLOWERING TIME. March to April. ETYMOLOGY. Frappier, in his description, refers to the nectar-like droplets on the disk of the lip. A viscous substance on the lip is not unique to this species and can be found in others, such as Liparis listeroides, but the distinct droplets are unusual. NOTES. Liparis nectarina first appeared in Frappier's list in 1880, without a description. It was then formally described by Cordemoy (1895) which incorporated Frappier's text on the Orchidaceae. Frappier did not indicate a type but gave the locality of 'Tampon Saint-Pierre'. It is assumed that Frappier's herbarium was lost  and there are no plants in Cordemoy's herbarium that are labelled as such or fit the description. It was therefore decided to assign a new type from a more recent and well-documented collection (neotype Bernet in Hermans 8153). It originated from a similar area in the South-central mountains of Réunion. There are Liparis nephrocardia Schltr. (Schlechter 1924: 145);Perrier (1936: 253;1939: 288); Hermans et al. (2007: 220); Cribb & Hermans (2009: 154); Bosser & Lecoufle (2011: 409); Hervouet (2018: 414). Types: Madagascar, Mt Tsaratanana, Jan. 1923, Perrier 15746 (P00095479) (lectotype P, designated here); Mt Tsaratanana, Dec. 1922, Perrier 15246 (P00095476) paratype P).
Schlechter compared Liparis nephrocardia with L. jumelleana but it is very different in habit and in having a somewhat longer and kidney-shaped blade of the lip. Liparis bosseri described above is similar in plant and flower size but the inflorescence of L. bosseri far surpasses the leaves (vs about the same length) and lacks the distinct rounded callus. It is similar to L. andringitrana in both habit and flower but L. nephrocardia is a little smaller and squatter, its lip is sub-rectangular near the base and the callus is lobed and much more obvious vs obscure, it also has more prominent column wings (Table 2). DISTRIBUTION. Endemic to Madagascar: Antsiranana, Fianarantsoa and Toamasina provinces only (Map 33). SPECIMENS EXAMINED. MADAGASCAR. Mt Tsaratanana, Jan. 1923, Perrier 15746 (P00095479) lectotype (P); Mt Tsaratanana, Dec. 1922, Perrier 15246 (P00095476), paratype (P); Sambirano, Northern slopes of the Sambirano valley, 600 m, Dec. 1922, Perrier 15745 (P); East, near the confluence of the Onive and the Mangoro, c. 700 m, Feb. 1925, Perrier 17135 (P); Mts N of Mangindrano, Ambohimirahavavy, 1800 m, Jan. 1951, Humbert 24887 (P); Fianarantsoa, Ranomafana area, Jan. 1992, Hermans 8286 (K). HABITAT. Humid evergreen forest and moss-forest. Altitude: 600 -1800 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis nephrocardia is estimated to be 20,007 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 16 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis nephrocardia is known from four subpopulations representing four locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. December to March. ETYMOLOGY. Refers to the kidney / heart shape of the blade of the lip. NOTES. This is one of the few species where Schlechter indicated a holotype specimen (Perrier 15746) '…to avoid future confusion' but, even then, there are two sheets with the same number, collected at the same locality but at a different time, the better specimen is designated here as the lectotype. Perrier (1936: 253) mentioned that Perrier 17135 differs a little from the holotype in that it has 3-veined sepals (not 5-veined), petals shorter than the dorsal sepal and with two veins only, a smaller lip and the callus which is narrowly 'V' shaped. But there is some variability in all the specimens recorded by Schlechter. There seem to be two distinct forms of this species, a smaller one from Mt Tsaratanana and the Sambirano region and a lusher lower altitude form from the Eastern forests where the plant is almost double the size and the inflorescence much longer. ILLUSTRATIONS. Figs 43, 44;Bosser & Lecoufle (2011: 409)   Liparis hildebrandtiana Schltr. (Schlechter 1924: 140); Perrier (1936: 251;1939: 286). Type: Madagascar, Mt Tsiafajavona, Perrier 13511 (holotype P).
Large, erect, terrestrial, epiphytic or rarely lithophytic plant, up to 28 cm tall but generally between 15 -20 cm, rhizome very short, repent, roots filiform, flexuose more or less pilose. Pseudobulbs stem-like, narrowly cylindricalovoid to subcylindrical but bulbous at the base, up to 6 cm long, 2 cm wide, covered by 2 -3 long and overlapping stem-sheaths, sometimes the older leafless pseudobulb obliquely divergent, with up to 5 leaves starting half-way up the stem and ending with 2 -3 larger ones towards the apex. Leaves erectly spreading, obliquely elliptic to ovateelliptic, apiculate to obtuse, base rounded-cuneate and then narrowed into a 1.5 -4 cm petiole, overall 7 -15 × 3 -5 cm, plicate, margins a little undulate to crenulate. Inflorescence erect, up to 15 cm, 2 -3 mm in diam., laxly with up to 11 flowers but generally fewer. Peduncle costate, with 2 -4 peduncle sheaths, the basal ones amplectant, the higher spreading, up to 35 × 4 mm. Rachis up to 8 cm long. Floral bracts erectly-spreading, variable but generally the lower ones lanceolate-acuminate, the higher ones shorter, ovate-acuminate to cordate and almost leaf-like, 6 -14 × 2.5 -5 mm. Flowers large, erectly spreading, up to 20 × 15 mm, sepals and petals greenish-white to more or less greenish yellow becoming more yellow-orange with age, the lip a little darker and sometimes with a darker patch on the disk, column white, anther white to greenish-white, pollinia yellow. Pedicel and ovary slightly ridged to costate, 9 -14 × 2 -5 mm. Dorsal sepal arching over the column, narrowly-triangular, base cordate, tip obtuse to acuminate, 12 -17 × 2 -3.5 mm. Lateral sepals porrect below the lip, frequently fused or part fused with one another, subfalcate, oblong-lanceolate, somewhat obtuse, 11.5 -16 × 3.5 -5 mm. Petals pendent and recurved towards the tip, narrowly linear narrowing towards the apex, margins incurved, 10 -15 × 0.5 -1.2 mm. Lip elliptic, apex truncate-obtuse and shortly indented, anterior margins undulate to dentate, strongly curved just above the middle, base shortly cordate-auriculate, with two distinct parallel calli at the base, more or less longitudinally semi-obovate to lobular, overall 9.6 -14.2 × 5 -7.4 mm, generally about half as wide as long. Column slightly curved, semi-terete, the margins extended into rounded wings along the apical half, 4 -7 × 1.5 -2.5 mm. Anther shallow, broadly obovate with a tiny apicule at the anterior margin, 1.0 -1.2 × 1.1 -1.3 mm. Pollinia ovate, c. 0.5 × 0.6 mm. Seed capsule costate. RECOGNITION. This is a variable species both in habit and flower but there are a number of distinguishing characteristics. It is a large plant with three to five broad ovate-elliptic leaves along a bulbous stem, a costate peduncle, broad, almost leaf-like floral bracts, large flowers with a long narrowly triangular dorsal sepal, a strongly curved lip almost twice as long as wide with a denticulate anterior margin, with parallel calli at the base and a rounded anther.
There are some strong similarities with Liparis sambiranoensis but that has a longer pseudobulb and a more distinct rounded epichile and angular hypochile of the lip (vs almost pandurate), a curved column with small angular wings make the latter species also distinct. Similar to L. ornithorrhynchos but the latter is generally shorter in growth, has a more rounded leaf, somewhat smaller flowers with narrower bracts, a narrower dorsal sepal, much rounder lip without distinct calli and an anther without a distinct beak. A number of Hildebrandt specimens from the same locality were confused with L. ornithorrhynchos. DISTRIBUTION. Widespread in Madagascar in the highlands and Eastern forests. Antananarivo, Antsiranana, Fianarantsoa and Toamasina provinces (Map 34). There is a record from Toliara (Rakotoson 9421) and they seem to be more compact elongate plants growing in a drier environment. Herbarium material collected by Boivin in 1850 -51 (P00095480-2) on Nosy-Be in Northern Madagascar could be a misidentification, the plants mainly consist of damaged or fruiting specimens; the floral bracts are not typical for this species. The species is rare on Réunion and records are relatively modern (Bernet 2010b: 69  Category LC: the extent of occurrence (EOO) of Liparis ochracea is estimated to be 783,412 km 2 (far exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 132 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis ochracea is known from 24 subpopulations representing 29 locations (sensu IUCN), this species has been preliminarily assessed as LC using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction.
Previously assessed and published (as Liparis hildebrandtiana) in the IUCN Red data listing for Réunion as CR (Picot 2013: 11). FLOWERING TIME. January to April. ETYMOLOGY. The name presumably refers to the ochre colour of the flower, especially the ochre area in the lamina of the lip mentioned by Ridley in his description. This is not often seen in this species; Ridley's description relied on Deans Cowan's fairly rudimentary watercolour which shows this feature. NOTES. The species was described in 1885 by Henry Ridley based upon a watercolour drawing in William Deans Cowan's sketchbook of Madagascan orchids (Cowan 1880: 17), now kept at the Natural History Museum, London. Ridley confirmed in his description that there was no herbarium specimen of it. Deans Cowan was not one of the most accomplished botanical artists but the watercolour shows the plant and also some detail of the flower and its parts; together with Ridley's description, the species is fairly well defined. In the same publication Ridley also described Liparis connata which he admitted is likely to be a monstrous form of another species where the lateral sepals are fused together; otherwise it shows all the characteristics of L. ochracea. In 1924 Rudolf Schlechter described L. hildebrandtiana; 'named in memory of the Madagascar researcher Johann Maria Hildebrandt', it has all the characteristics of L. ochracea. Perrier (1936: 242, 251;1939: 286) reduced both L. ochracea and L. connata to the synonymy of L. hildebrandtiana: in his reasoning he mentioned that he thought L. hildebrandtiana is nothing else but the Liparis named by Ridley as L. ochracea and L. connata but that it was impossible to verify this synonymy as L. ochracea was only described from a watercolour by Deans Cowan and L. connata is an abnormal cleistogamous form. He proposed therefore to conserve, notwithstanding its later publication, the name of Schlechter which, he said, is at least applied to a clearly defined species and is represented by several herbarium specimens, showing clearly the variability of the species and the cleistogamous forms that exist. This statement may well be true but is not valid under the rules of nomenclature and therefore the name L. ochracea has priority under ICN article 11. 4. Both Ridley and Perrier mentioned the great variability of the species plus the tendency to produce cleistogamous or deformed flowers, this is also confirmed by observations in the field. Perrier (1936: 251) gave a detailed analysis of the different variants and abnormalities found in plants from different localities, he also annotated a number of herbarium specimens; fusion of the lateral sepals seems to be quite common and the absence or deformity of anther and stigma are also frequent, especially the flowers at the apex of the inflorescence. He mentioned that the shape of the lip can also vary but the number of veins (seven) and their disposition (three median undivided and four branched laterals) are always present.
A specimen in the Paris herbarium (Boivin 1035 P00409214) from Réunion, misidentified as Liparis caulescens, is annotated as L. polycardia Rchb.f. but is in fact L. ochracea. ILLUSTRATIONS.  Small erect to arching terrestrial plant on a short to repent rhizome, rhizome up to 5 cm long covered with thin sheaths, roots filiform more or less villous 3 -5 mm diam., overall up to 15 cm but often smaller. Pseudobulbs stem-like, up to 5 cm long, 7 -10 mm diam., completely covered by 2 -3 papery sheaths up to 45 mm long, with 2 or rarely 3 leaves at the top, disintegrating when the new growth emerges. Leaves shortly petiolate, ovate, acute, somewhat plicate, 3.5 -7.5 × 1.5 -3.5 cm, green. Inflorescence erect, laxly 8 -12-flowered, overall between 6 and 12 mm long, 3 -4 mm diam. Peduncle angular, with 2 -3 acutely lanceolate peduncle sheaths, 8 -9 × 3 -4 mm. Rachis between 4 and 6 cm. Floral bracts ovate-lanceolate, the upper ones more acutely-lanceolate, 4.2 -11.5 × 1.8 -2.9 mm, decreasing in size towards the apex. Flowers large, spreading, up to 16 × 10 mm, overall pale green becoming more yellowish-green with age, the column white, with a darker green longitudinal band on the disk of the lip. Pedicel and ovary long and thin, up to 15 × 1.5 mm. Dorsal sepal linear, erect to arching, margins incurved, 8.5 -11.1 mm. Lateral sepals ovate-lanceolate, obtuse, folded underneath the lip, 9.3 -9.5 × 3.8 -5.2 mm. Petals narrowly linear, spreading, 10.8 -10.6 × 0.6 -1 mm. Lip rounded to cordate with a short, narrow auriculate base, margins slightly undulate, a short apicule at the tip, with a broad disk and without calli, 7.9 -9.2 × 8.0 -8.3 mm. Column arching, almost straight, with long elongate wings, up to 5 mm long, 1.5 mm wide. Anther ovate with an elongate-rostrate obtuse beak, 0.9 -1.0 × 1.0 -1.2 mm. Pollinia ovate, c. 0.5 mm. Seed capsule obovate, slightly ridged, up to 15 × 8 mm. RECOGNITION. The plant has a creeping rhizome, a short thin stem covered with two to three long papery sheaths, two ovate, plicate leaves, ovate-lanceolate floral bracts becoming narrower and shorter towards the apex of the inflorescence, a long and thin pedicellate ovary, and relatively large flowers, with the lateral sepals folded beneath the lip and about twice as long as wide, a roundly to broadly cordate lip with a narrow base, no calli and a distinctly beaked anther cap.
In his description Ridley compared it with Liparis bowkeri from mainland Africa but that has a different habit, rounded vs stem-like pseudobulbs and also has two distinct calli at the base of the lip. In habit and lipshape L. ornithorrhynchos is similar to both L. imerinensis and L. andringitrana but it does not have the flattened pseudobulbs of L. andringitrana, the flowers almost 1 =3 larger and lack a callus at the base of the lip. It is often confused in herbaria with L. ochracea but its habit, smaller and rounder lip lacking a callus and its beaked anther cap make it distinct. The syntype or lectoparatype (Hildebrandt 4049) at K differs in its lip shape from that at W and elsewhere; the collection is a mixed group containing L. ochracea. DISTRIBUTION HABITAT. Highland evergreen forest, in shade, damp places. Altitude: 1000 -1500 m. FLOWERING TIME. January to April. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis ornithorrhynchos is estimated to be 11,163 km 2 (which exceeds the limits for Endangered status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 16 km 2 (which also falls within the limits for Endangered status under the criterion B2). Liparis ornithorrhynchos is known from four subpopulations representing four locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. ETYMOLOGY. Presumably referring to the resemblance of the anther to the beak of a bird. NOTES. Ridley (1885) described Liparis orthorrhynchos and cited two distinct collections: material collected by the Rev. William Deans Cowan in 1880 and specimens collected in 1881 by Johann Hildebrandt from a similar area. The Deans Cowan material consists of a single herbarium sheet of several plants kept at BM, it is annotated '35' in Ridley's hand which refers to one of Dean's Cowans watercolours (Cowan 1880: 17, no 35). Hildebrandt's collection is represented in several European herbaria. The Deans Cowan specimen was selected as the holotype because it corresponds well to Ridley's description and also matches Dean's Cowan's illustration. The Hildebrandt material is a mixed group containing some atypical plants that do not correspond well with the description, the majority of plants being in seed. They are considered to be syntypes or lectoparatypes. ILLUSTRATIONS. Fig. 46;Cowan (1880: 17 no 35); Cribb & Hermans (2009: 157) is Liparis ochracea.
The species has some similarities with Liparis cladophyllax and L. dryadum but neither have the distinct globular pseudobulbs and have a column that is indistinctly winged vs distinctly angular, and a lip almost half the size. Liparis parva bears some similarity to L. cespitosa but that species has a single leaf and flowers half the size (Table 1). DISTRIBUTION. Endemic to Madagascar: Fianarantsoa and Toamasina provinces (Map 36). Szelengowicz et al. (2011: 11) and Szelengowicz & Tamon (2013: 367), cited and illustrated this species from Réunion: it is not possible to verify the identity of the plant from the photographs but it is unlikely to be this species. SPECIMENS EXAMINED. MADAGASCAR loc., May 2018, Hermans 8218 (K). HABITAT. In wet Eastern and highland evergreen forests, in deep shade, on moss and lichen-covered trees. Altitude: 1100 -1500 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Liparis parva is estimated to be 7,631 km 2 (which exceeds the limits for Endangered status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 12 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). With only three known subpopulations representing three locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction.
Onive & Mangoro, terrestrial, 700 m, Feb. 1925, Perrier 17136 (holotype P (P00095499); isotype (P90005500) P). Liparis perrieri Schltr. var. trinervia H.Perrier (1936: 251) 3.8 -5.8 × 0.4 -0.9 mm. Lip ovate to flabellate-dilate, the base rectangular strongly auriculate, anterior margin recurved, retuse, more or less denticulate, carrying 2 more or less prominent horn-like calli at the very base, 3.1 -4.5 × 3 -4.6 mm. Column curved forward in the upper half, strongly roundly winged towards the apex, 2 -3.2 × 0.8 -1.2 mm. Anther obovoid, often with a long apicule at the tip, 0.8 -1.1 × 0.7 -0.9 mm. Pollinia ovoid 0.3 -0.5 × 0.5 -0.7 mm. Seed capsule narrowly obovate, up to 12 × 5 mm. RECOGNITION. This is a large, somewhat variable species with one flowering growth with many small flowers and the remnants of the previous year's growth. It has a ridged peduncle, strongly decurved floral bracts, green flowers tinted with purple, a much darker violet to purple lip with two separate horn-like calli at the very base, calli that are easily missed in dissection of herbarium material, a column winged near the apex and an obovoid anther without a distinct apicule. The species is similar to Liparis nervosa and its numerous variants which has a widespread distribu-HABITAT. The species is mainly restricted to the Western forest where it is found in the more humid areas growing on granite, calcareous rock or the base of trees on humus in moderate shade. Also in evergreen humid forest. Altitude: 500 -1200 m.
CONSERVATION STATUS. Category LC: the extent of occurrence (EOO) of Liparis perrieri is estimated to be 252,553 km 2 (far exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 76 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis perrieri is known from 18 subpopulations representing 18 locations (sensu IUCN), this species has been preliminarily assessed as LC using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. December to February. ETYMOLOGY. Named for Henri Perrier de la Bâthie (1873-1958. NOTES. It was described and illustrated by Schlechter (1913) based on several specimens collected by Perrier. In the description, he referred to Perrier 479 as the type, this is a typographical error and it should be Perrier 499 which is from the correct locality and on the type sheet in P confirmed by Perrier and Schlechter. The type consists of two sheets and sheet P00095497 is designated as the lectotype because it is Map 37. Distribution of Liparis perrieri. more complete, P00095498 is an isolectotype. Schlechter also mentioned Perrier 45 (1886 of Perrier) and this is labelled as a 'co-type 'or syntype. Liparis perrieri is a variable and widespread species and exhibits a considerable range in plant and flower size and shape . Perrier (1936: 251) described var. trinervia without a Latin diagnosis but the name was later validated by Hermans (in Hermans et al. 2007: 289). It is said to differ from the typical form by the "lateral sepals with five veins; the lip being shortly apiculate in the middle of the front indentation, with three veins, the middle one simple, and with a well separated callus, in a conical swelling and very pronounced; the anther with the front lip subrectangular". It comes from the Eastern side of the island and is a local variant within the morphological range of the species. Although it is said to differ in having lateral sepals with five veins this is not consistent and can also be found in the syntype of the typical variety (Perrier 1886), the different lip, callus and anther shape can also be observed in other specimens of the typical Western form.
Liparis polycardia Rchb.f. (Reichenbach f. 1885: 543 Lip obcordate to cuneate to expanded flabellate, anterior margin slightly undulate to almost dentate, base hardly winged with two distinct, more or less flattened horn-like calli at the base, 5.1 -6.9 × 4.2 -6.6 mm. Column slightly curved at the apex a little thickened at the apex and then with two small but distinct triangular wings, 3.5 -4 × 0.7 -1.1 mm. Anther more or less sharply beaked, the pollen chambers only slightly divided, 1 -1.3 × 0.7 -0.9 mm.

RECOGNITION.
A large plant with a very long inflorescence and small to medium-sized flowers, the stem-like pseudobulbs covered with long attenuate, papery sheaths, longly ligulate and attenuate leaves, an obcordate lip lacking wings at the base and with two sharplyhorned calli at the base, and a strongly beaked anther.
In his description Reichenbach mentioned that the flowers are like those of Liparis foliosa, referring, as explained above, to L. reflexa (R.Br.) Lindl. (Lindley 1825: t.882) from Australia, but plant habit, length of the inflorescence and the shape of the lip are different with the leaves about half the size, the flowers a little smaller with the lip obcordate vs oblong and the callus horn-like vs indistinct. It shares the basal teeth of the lip with L. listeroides and L. imerinensis but the plant and flower are bigger, the leaves longer and narrower (ligulate vs oval-elliptic) and the inflorescence longer by a third.
The most closely allied species is Liparis stenophylla which is similar in habit, flower size and shape but its leaves are more horizontal rather than erectly spreading, the leaves are narrowly lanceolate (vs lanceolate in L. polycardia), the petiole shorter, the inflorescence generally shorter, the petals shorter and the lip obovate to suborbiculate rather than obcordate.
Some forms have larger than average flowers and a less distinctly heart-shaped lip with a more denticulate anterior margin (Hermans 6737 and Antilahimena 4753), the drawing by Reichenbach accompanying the type also shows a denticulate lip. DISTRIBUTION. Endemic to Madagascar: Antsiranana,Fianarantsoa and Toamasina provinces (Map 38). The label on the type specimen in W indicates Comores as a locality but this has been corrected to Madagascar which corresponds with Humblot's travels as shown in his notebooks (Humblot 1882(Humblot -1885. The label indicated 'Forêt d'Anhaya' as the collecting locality but no reference has been found to this name. A specimen in the Paris herbarium (Boivin 1035 minimal area of occupancy (AOO) is estimated to be 48 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis polycardia is known from nine subpopulations representing 10 locations (sensu IUCN), this species has been preliminarily assessed as VU using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. February to June. ETYMOLOGY. Probably referring to the many heartshaped lips on the flowers. NOTES. Liparis polycardia has remained somewhat enigmatic for some time . Reichenbach f. (1885) validly described it, together with a number of other new species discovered by Leon Humblot. It was included in Ridley's 1887 monograph of the genus where Reichenbach's description was repeated and marked as 'unseen' and possibly related to L. puncticulata, and said to have come from SE Africa. Kuntze (1891: 671) transferred it to his new genus Leptorchis together with other Liparis. It is not included in subsequent litera-ture except for a listing in Durand & Schinz (1895: 9) and repeated by Voeltzkow (1917: 444) as coming from Africa. It is recorded in Govaerts (2017) as an unplaced name. ILLUSTRATIONS. Figs 51, 52 and 53. Liparis puncticulata Ridl. (Ridley 1886a: 119;1886b: 270);Perrier (1936: 245;1939: 266); Hermans et al. (2007: 222); Cribb & Hermans (2009: 150); Hervouet (2018: 461). Types: Madagascar, Baron 4334 (K000242144) lectotype K, designated here; isolectotypes Baron 4334 BM (BM00090071), P (P0095501)). Leptorkis puncticulata (Ridl.) Kuntze (1891: 671).
Large slender, erect plant, terrestrial or rarely epiphytic up to 40 cm tall, often forming clumps of 5 or more growths, on a short 0.3 -2 cm rhizome, roots flexuose, woolly. Pseudobulbs slender, stem-like, hardly thickened at the base, 16 -38 cm long, 5 -9 mm diam. Covered by 6 -9 membranous sheaths slightly overlapping in the lower half and becoming more distant towards the apex, keeled and acuminate at the tip, characteristically spotted with more or less visible dark dots, with 3 -4 somewhat distant leaves at the apex. Leaves narrowly lanceolate to ligulate, on a short 3 -8 mm petiole, acute, margins generally a little undulate, sometimes irregularly serrate, 5.5 -8 × 1.2 -2.5 cm, pale green. Inflorescence erect, thin c. 2 -3 mm diam., up to 30 cm tall but generally shorter, carrying up to 11 flowers but most often far fewer. Peduncle somewhat ridged, with a few 8 -14 × 3 -4 mm sterile bract-like sheaths. Rachis up to 18 cm, laxly racemose, the lower flowers fading or disappearing well before the apical ones open. Floral bracts erectly spreading, lanceolate acute 7.2 -11 × 2 -3 mm. Flowers medium size, up to 19 × 12 mm, erectly spreading, pedicellate ovary pale greenish white becoming darker towards the base of the flower, sepals and petals pale to olive green, lip olive green, the basal lobes paler, with a characteristic very dark green disk, column white, anther greenish-yellow, all parts becoming yellow to pale orange with age. Pedicel and ovary slender erect, 10 -21 × 1.1 -1.8 mm. Dorsal sepal lanceolate to ligulate-angustate, the base strongly lobed and for the most part strongly incurved, 7.7 -14.5 × 1 -1.8 mm. Lateral sepals broadly lanceolate, obtuse, curved alongside or under the lip, 6 -12.5 × 3 -4 mm. Petals oblanceolate to linear, strongly incurved along their length, 9 -15 × 0.5 -1 mm. Lip cuneate, the base thickened, longly auriculate, at first hollowed into a narrow gutter then widened into an obtriangular blade, bilobed to obscurely obcordate at the anterior margin which is distinctly dentate. Without a distinct callus at the base but with a thickened bilobed ridge with a rhomboid depression in-between, overall 8 -13 × 4 -6.7 mm. Column suberect, the tip curved, wings short and obscure, 3.5 -6 × 0.6 -1.5 HABITAT. Terrestrial in humus and in moss-rich forest or on the base of moss-covered trees. Altitude: 800 -2500 m. CONSERVATION STATUS. Category VU: the extent of occurrence (EOO) of Liparis puncticulata is estimated to be 101,847 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 64 km 2 (which falls within the limits for Endangered status under the criterion B2). Liparis puncticulata is known from 14 subpopulations representing 14 locations (sensu IUCN), this species has been preliminarily assessed as VU using the green listing method. This species is threatened by mining activities, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. The main concentration of colonies is in the Mandraka area which has now been greatly eroded and only few forest remnants remain. FLOWERING TIME. January to July but mainly January and February. ETYMOLOGY. Refers to the black spots, more or less visible, on the stem sheaths. VERNACULAR NAMES. Avonala (Herb. Jard. Bot. Tan. 3670 in P). NOTES. Three herbarium sheets of Baron's type material exist, one in BM which has no flowers, one at P with flower fragments only and one at K which has several good flowers; for this reason the K specimen has been designated as the lectotype, the others are isolectotypes. There also is a tracing by Reichenbach of the Kew type at W. ILLUSTRATIONS. Figs 54, 55;Perrier (1939: 271); Hermans et al. (2007: pl.46); Cribb & Hermans (2009: 151); Guérin & Hervouet (2011: 96);Hermans (2013: 293); Hervouet (2018: 416 Very small epiphytic or terrestrial plant 5 -7 cm high, on a short rhizome, roots wiry, woolly, c. 3 mm diam. Pseudobulbs ovoid to conical 12 -15 × 9 -10 mm, the base covered by membranous sheaths without strong veins, carrying 2 -3 erectly-spreading leaves. Leaves ovate to lanceolate, acute, narrowed into a short petiole at the base, blade corrugate, 3.8 -6 × 1.5 -2.1 cm, pale green. Inflorescence erect, up to 95 mm, corrugate, with up to 13 flowers. Peduncle with 2 -3 small narrowly lanceolate bracts. Rachis loosely racemose, in the upper third of the inflorescence. Floral bracts narrowly lanceolate 3.5 -5.5 mm, about 1 3 the length of the ovary, minutely erose at the base. Flowers small, erectly spreading, av. 9 × 7 mm, floral segments overall pale and translucent, with minute dark spots within the structure, especially the lip (not always obvious in dried specimens). Pedicellate ovary green, perianth yellowish green, becoming yellow-orange with age, lip and column green, anther white-brown. Pedicel and ovary cylindrical corrugate, 8 -13 × 0.7 -0.8 mm. Dorsal sepal erect, linear-ligulate, more or less widened at the base, margins recurved, 6.8 -7 × 1.1 -2 mm. Lateral sepals broadly oblong-falcate, obtuse, curved below the lip and overlapping, spreading,margins incurved, Lip transversally elliptic to reniform, with distinct short wings at the base then abruptly broadened into the blade which is slightly undulate at the anterior margin and minutely mucronate at the apex, two distinct calli (appearing as an angular cushion in living material) at the base ending in longitudinal rounded ridges along the central vein, hunchbacked,broadened   RECOGNITION. This is a very small plant, with ovoidconical pseudobulbs, ovate, corrugate leaves, and small flowers with translucent and punctate segments, a transversally elliptic to reniform lip wider than long (Frappier, in his key, referred to the lip being orbicular but this may have been a misinterpretation especially as he also described it as very obtuse, all other characteristics correspond with his description), shortly auriculate at the base, mucronate at the tip, and with an angular bilobed callus at the base, a short column and thickened at the apex, wings long and indistinct, and an anther with a small lobule in front.
The plant has some similarities with Liparis bathiei, L. flavescens and L. lutea but their inflorescences are denser, their lip ovate (vs transversally elliptic), their column wings more angular and their anther with a more acute beak. DISTRIBUTION Category EN: the extent of occurrence (EOO) of Liparis punctilabris is estimated to be 169 km 2 (which falls within the limits for Endangered status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 20 km 2 (which also falls within the limits for Endangered status under the criterion B2). With only three known subpopulations representing two locations (sensu IUCN), this species has been preliminarily assessed as EN using the green listing method. This species is threatened by grazing and anthropogenic fires, resulting in habitat reduction and habitat quality reduction.
Previously assessed and published in the IUCN Red data listing as EN (Picot 2013: 12). FLOWERING TIME. January to April. ETYMOLOGY. Referring to the punctate lip (mainly visible in living flowers).

NOTES.
Liparis punctilabris first appears in Frappier's listing of Réunion plants in 1880 but without a description. It was then described by Cordemoy (1895) who incorporated Frappier's manuscript.
Frappier indicated several localities in Réunion but there is no clear type material. A fruiting specimen is in the Cordemoy herbarium (MARS) corresponds with the description and comes from Piton Hyacinthe which is one of the localities mentioned by Frappier in Cordemoy. There is no direct proof that this specimen is part of the type material and the fruiting plant is a poor representation of the species. As there is no other historical representative material available it was decided to assign a new type from a more recent and well-documented collection (Bernet in Hermans 8161). The new type originates from a similar area in the South-central mountains of the island. ILLUSTRATIONS. Figs 56, 57;Bernet (2010a: 96); Szelengowicz & Tamon (2013: 368).
Liparis rivalis Schltr. (Schlechter 1924: 146);Perrier (1939: 267); Hermans et al. (2007: 223) ;Cribb & Hermans (2009: 150); Hervouet (2018: 417 Large erect terrestrial plant, up to 35 cm tall; rhizome very short; roots more or less pilose, c. 1 mm diam. Pseudobulbs slender, stem-like, up to 22 cm × 3 -6 mm, partly covered by 4 -8 amplectant sheaths, sometimes with a leaf-like bract just below the leaves, with 3 -5 apical sub-petiolate leaves. Leaves asymmetric, somewhat falciform, thin, lanceolate to elliptical-lanceolate, acuminate, up to 5 -15 -2 -5 cm but usually in the smaller range, pale green. Inflorescence erect, longer than the leaves, carrying up to 30 flowers, up to 15 cm. Peduncle with a few short costate peduncle sheaths sizes and several sterile bracts. Rachis densely-flowered, racemose, up to 14 × 2 cm diam. Floral bracts spreading, linear-lanceolate, acuminate, the lower ones a little longer than the pedicellate ovary, the apical ones shorter, 5.2 -7 × 0.9 -1.6 mm, green. Flowers small, up to 9 × 6 mm, erectly-spreading, lower ones fading before the upper ones open, ovary and column green, sepals and petals brownish-green, lip pale orange and becoming darker towards the middle  Antsiranana, near Mt d'Ambre, Jan. 2017, Hermans 8106 (K). HABITAT. Terrestrial at the edge of streams or damp places, in undergrowth in evergreen forest of gneiss and laterite. Often in deep leaf litter and moss. Altitude: 800 -1600 m. CONSERVATION STATUS. Category VU: the extent of occurrence (EOO) of Liparis rivalis is estimated to be 40,636 km 2 (exceeding the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 32 km 2 (which falls within the limits for Endangered status under the criterion B2). With only eight known subpopulations representing eight locations (sensu IUCN), L. rivalis has been preliminarily assessed as VU using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. December to March. ETYMOLOGY. Refers to the habitat, near the margin of streams, where the type specimen was found. NOTES. Rudolf Schlechter (1924: 146) described Liparis rivalis from material collected by Perrier on Mt Tsaratanana. There are two sheets of type material: lanceolate peduncle sheaths c. 5 × 2 mm. Rachis subcorymbose, elongating with age, with up to 16 flowers, generally fewer. Floral bracts lanceolate, 5 -5.8 × 1.2 -1.5 mm, much shorter than the pedicellate ovary, reddish-brown. Flowers medium in size, up to 1.4 × 1.5 mm, ovary, dorsal sepal and petals burgundy, lateral sepals yellow-brown more or less marked with burgundy, lip pale-yellow with the veins dark burgundy, column white marked with purple, anther pale yellow, pollinia orange to brown. Pedicel and ovary tubular to fusiform, twisted, sulcate, up to 12 × 1.5 mm, elongating with age. Dorsal sepal recurved, subulate to linear, margins incurved, subapiculate, 5.5 -8 × 0.7 -1.4 mm. Lateral sepals oblong ovate, obtuse, connate at the base, folded underneath the lip, 4.7 -6.7 × 2 × 2.9 mm. Petals descending to divergent, subulate to linear, the margins more or less incurved, subapiculate, obtuse, 4.5 -5.6 × 0.2 -0.3 mm. Lip suborbicular to obcordate-triangular, curved downwards toward the middle, base with long upturned lobes entire at the margins, forming small auricles at the base, emarginated to apiculate at the front, anterior margins serrate to laciniate, with a more or less pro-nounced basal callus which can be divided, 3.5 -5.9 × 4.2 -4.9 mm. Column elongate, arcuate toward the apex, wings at the tip obtusely triangular, 2 -3.4 × 0.7 -1.5 mm. Anther ovoid, strongly beaked, c. 1.1 × 0.9 mm. Pollinia ovoid c. 0.4 × 0.3 mm. Seed capsule oblong, a little narrowed from the apex to the base, pedicel slender, c. 12 × 3.5 mm. RECOGNITION. A distinctive small plant with a short thin pseudobulb and two to three, cordate to ovate, purplered leaves, with undulate margins, at the top. Its inflorescence is subcorymbose and bears medium-sized, more or less burgundy coloured flowers, with a paler lip with a serrate anterior margin and indistinct rounded callus. The anther is distinctly beaked. It is quite distinct from all the other Madagascan and Mascarene species in its typical habit, lip shape and colour. DISTRIBUTION Large erect terrestrial, lithophyte or rarely epiphytic plant 15 -40 cm high, emerging directly from the base of the previous growth, roots filiform, more or less villous. Pseudobulbs cylindrical, narrowly elongate, sometimes somewhat flattened-angular, almost entirely covered by membranous sheaths and by the sheathlike petiole of the lower leaves, with 3 or 4 to rarely 5 leaves towards the top, growth emerging from the previous leafless growth which is divaricate, older pseudobulbs disintegrating and merging in the substrate, without obvious rhizome, 7 -18 × 2 -3 cm. Leaves erectly spreading, variable in size, often in two pairs 15 -30 mm apart, blade elliptic-acuminate, corrugate, narrowed into a long petiole amplectant to the pseudobulb, overall 10 -18.5 × 5 -7.5 cm, pale green, more or less glossy. Inflorescence erect, up to 30 cm long, from the centre of the growth, much longer than the leaves, carrying up to 25 flowers, the lower ones fading before the upper ones open. Peduncle costate, with a few ovate-lanceolate sterile bracts. Rachis laxly racemose, up to 15 cm. Floral bracts shorter than the pedicellate ovary, porrect, ovatelanceolate, strongly attenuate at the tip, almost cordate at the base, 8 -12 × 3.8 -4.1 mm, green. Flowers large, erectly spreading, up to 29 × 14 mm, ovary pale green, sepals and petals pale greenishyellow, lip darker olive green to orange, column and anther white, all parts becoming paler and more yellow with age, tepals and lip a little glossy. Pedicel and ovary porrect, slightly upturned toward the base of the flower, cylindrical, somewhat grooved, 10 -17 × 1.5 -2 mm. Dorsal sepal erect to incurved, linearlanceolate, base cordate-auriculate, margins more or less recurved, 12.5 -18 × 1.2 -3.8 mm. Lateral sepals parallel, descending below the lip, obliquely oblong, obtuse, 11 -16 × 3.4 -5.9 mm. Petals divaricate, obliquely linear, margins recurved, obtuse, 11.4 -17 × 0.5 -1.5 mm. Lip strongly curved in the middle forming a gutter-like hypochile, shortly auriculate at the base and a broadly oval hypochile serrate to crenate at the anterior margin, with a bi-gibbose callus at the base, 9.1 -14 × 5.4 -8.8 mm. Column strongly curved towards the apex, with distinct sharply triangular wings, 4.1 -6 × 1.1 -1.5 mm. Anther ovoid, with a short acute beak at the tip, 1.1 -1.5 × 0.9 -1.0 mm. Pollinia ovoid c. 0.5 × 0.7 mm. RECOGNITION. This is a large plant, up to 40 cm tall, with a distinct thin stem and about four glossy leaves borne a few centimetres apart, old pseudobulb divaricate, large flowers in a many-flowered lax raceme. Its lip is longer than wide, narrowed around the middle, serrate or crenate on the front margin and with a bigibbose callus at the base, and its column has sharply triangular wings and a shortly beaked anther.
It is closest to Liparis ochracea but L. sambiranoensis has longer pseudobulbs, generally a longer inflorescence, a longer lip with a rounded epichile and angular hypochile (vs almost pandurate) and short angular vs elongate column wings. There are similarities with L. ornithorrhynchos but this species has a different habit with shorter pseudobulbs, fewer flowers, the lip is almost pandurate vs rounded and the anther sharply triangular vs elongate-obtuse. It is also similar to L. bemarahensis described above but it Very small to small terrestrial or epiphytic plant, very variable in size, 3 -8 cm tall at most, few wiry roots, generally in small tufts on a short rhizome. Pseudobulbs completely or partly submerged in the soil or humus, broadly oval to subcylindrical-conical, single-or 2-leaved, very small, 3.1 -10 × 2.3 -8 mm, covered with 2 -3 thin white sheaths, one larger than the others, up to 10 -18 × 4 -5 mm. Leaves emerging from the apex of the pseudobulb, mainly carried on the flowering bulb, lanceolate, acute, margins sometime slightly undulate or minutely dentate, with a prominent central vein, narrowed into a short petiole at the base, somewhat fleshy, pale to dark green, with a prominent central vein, overall 15 -60 × 4 -15 mm. Inflorescence erect, wiry, much longer than the leaf, up to 12 cm long, with 3 -7 flowers towards the apex. Peduncle disproportionally long, generally well over half the length of the inflorescence, carrying an occasional peduncle scale 4 -6 mm long, sometimes with one or two sterile bracts. Rachis densely racemose, up to 2 -5 cm long. Floral bracts half the length of the ovary at most, lanceolate, acute, 2.1 -3.9 × 0.5 -1.3 mm. Flowers variable in size, very small to small, thin, green, yellowishgreen to orange-pale yellow, the petals often a little paler, the lip with a darker green glossy central groove, the callus also darker green, apex of the column white, anther white, 5 -10 × 4 -8 mm. Pedicel and ovary conical, slender, 4.5 -7.5 × 0.3 -1.1 mm. Dorsal sepal erect, lanceolate, margins incurved, more or less cordate at the base, 4.8 -7 × 1.2 -2.1 mm. Lateral sepals curved beneath the lip, oblong-falcate, 3.5 -5.1 × 1.8 -2.5 mm. Petals deflexed to spreading, linear, subulate, 4.4 -6.7 × 0.3 -0.4 mm, slightly spathulate at the apex. Lip broadly oval to suborbicular, rounded or slightly emarginated-auriculate at the base, anterior margin more or less crenulate, a hollow triangular groove runs from the base towards the anterior margin which is glossy and shiny in living plants, with a comparatively large bilobed, slightly depressed callus at the base, overall 2.8 -5.5 × 3.1 -5.5 mm. Column broadened at the base, acutely curved at the apex with the green listing method. In Madagascar, this species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming) while in Réunion island, it is threatened by grazing and anthropogenic fires, resulting in habitat reduction and habitat quality reduction Previously assessed and published in the IUCN Red data listing for Réunion as LC (Picot 2013: 17). FLOWERING TIME. January to May. ETYMOLOGY. Liparis scaposa is thought to refer to the long rachis, L. verrucosa is said by Frappier to refer to the uneven groove in the centre of the lip. Liparis microcharis refers to the small dainty appearance of the plant. NOTES. Liparis scaposa was first collected by Boivin during his 1847 -1852 expedition, on Réunion. Frappier found it in the Paris Herbarium and his notes were posthumously published by Cordemoy (1895: 183) who mentioned that there was some doubt about the identity of L. scaposa and its similarity to L. verrucosa which was described just two pages later; the main difference given in the key being the absence or presence of calli on the lip. There is no doubt that the two species are the same and have been recognised as such by recent authors (Bernet 2010a: 100). Frappier also mentioned two varieties: 'plana' and 'undulata' which are only different in the shape of the leaf margins and are well within the variability of the species. It was thought to be endemic to Réunion but close examination of living and herbarium material shows that there are a number of reliable records of this species from Madagascar which were previously identified as L. microcharis, now considered a synonym. Finet (1909: pl.1) described and illustrated it based on the Boivin type. Liparis microcharis was described by Schlechter (1924: 144) and based on a Perrier collection from S of Ambositra in central Madagascar; it shares all the characteristics of L. scaposa apart from its sometimes smaller flower size. Generally, the plants and flowers from Réunion are a little larger than the ones found in Madagascar but are well within the variability of the species. Szelengowicz et al. (2011: 11) reported L. microcharis from Réunion; the description and photographs confirm that this is L. scaposa.
The most closely allied species is Liparis polycardia which is similar in habit, flower size and shape but its   Inflorescence erect, far exceeding the leaf, 9 -16 cm long, 1 -2 mm in diam., carrying up to 6 flowers. Peduncle up to 10 cm, about double the length of the raceme with 2 -4 sterile bracts, 9 -11 × 3 -4 mm. Rachis loosely racemose, up to 7 cm long, the lower flowers fading before the apical ones open. Floral bracts narrowly lanceolate, a little cordate at the base, 6 -10 × 1. Lip longly-auriculate at the base, expanded into an elliptic blade angular at the base and dentate in the anterior margin, a short thickened central ridge at the base and slight swelling along the central vein, 15.2 -16.1 × 11 -11.4

mm.
Column relatively short and thick for the genus, slightly curved, with very broad thin rounded wings, 4.9 -6.1 × 2.5 -3.1 mm. Anther strongly sharply beaked, around twice as long as wide, 1.5 -2 × 0.8 -0.9 mm. Pollinia 2, ovoid, c. 5 × 3 mm. RECOGNITION. Liparis superclareae has tubular pseudobulbs, covered by brown sheaths, a single, lanceolate leaf, disintegrated on previous growth, and very large flowers, with an elliptic lip, dentate at the margins with a thickened ridge at the base, and a relatively short and broad, widely winged column, with a very long sharp beaked anther.
Liparis superclareae is close to L. clareae with which it shares the general habit and lip shape but the new species has much longer pseudobulbs, the leaves of the previous growth disintegrate before the new ones develop (vs persistent), its leaves are lanceolate (vs ovate) and are at least 2.5 -3 times longer than wide (vs 2 times), the floral bracts are less cordate at the base, the flowers and all its segments much larger, the column shorter with broad long wings (vs shortauriculate) and the anther much more sharply beaked. It is also similar to L. warpurii with which it shares the leaf and lip shape but the new species has a more linear plant habit (vs scrambling), the leaves of the previous growth disintegrate before the new ones develop (vs persistent), the floral bracts are less cordate at the base, the flowers and all its segments much larger, the lip has a thickened ridge (vs rounded callus), the column is short and broadly winged (vs shortly obtuse) and the anther much more beaked. It differs from L. magnifica and L. laurentii, described above, by its much larger size, shape of the lip, callus and anther, as shown in Table 8. DISTRIBUTION Category CR: The extent of occurrence (EOO) of Liparis superclareae cannot be estimated since it is only known from one subpopulation whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). Liparis superclareae is only known from one subpopulation representing one location (sensu IUCN), this rare species has thus been preliminarily assessed as CR using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. May to June. ETYMOLOGY. Referring to the much larger size of plant and flower compared with Liparis clareae. NOTES. Liparis superclareae is currently only known from the type specimen collected by Patrice Antilahimena in Maroantsetra. The collection consists of several flowering plants with type material in K, P and MO. The species overlaps in distribution with L. clareae which is more common in central Eastern forests. It could be considered a lower altitude, larger-flowered local form of this species but the pronounced deciduous habit, pseudobulb, leaf-shape and the shape of the column and anther make it distinct enough to be considered a distinct species. ILLUSTRATION. Fig. 65. the apex, 4 -6.1 × 2.3 -2.6 mm. Flowers erectly spreading, small, c. 14 × 8 mm, greenish-yellow. Pedicel and ovary glabrous, a little ridged, c. 9 × 1 mm. Dorsal sepal erect, ligulate-lanceolate, obtuse with the base shortly auriculate, 7.1 -8.1 × 0.9 -1 mm. Lateral sepals porrect, obliquely-oblong, obtuse, 6.8 -7.1 × 2.3 -2.5 mm. Petals narrowly linear, margins recurved, a little spathulate at the apex, 6.3 -6.5 × 0.4 -0.6 mm. Lip slightly curved, ovate-trullate, a little expandedangular towards the lower third, then slightly contracted towards the upper third and rounded at the top, a little dentate around the margins, base very slightly auriculate, on top with a small bipartite flattened callus, 6 -6.3 × 3.1 -3.3 mm. Column curved, wings indistinct, 3 × 0.8 mm. Anther ovoid, distinctly beaked, a little blunt at the tip, c. 1 × 0.6 mm. RECOGNITION. Liparis trulliformis is a small squat plant with elliptic plicate leaves, the new growth emerges from the disintegrating divergent or decumbent previous growth, an inflorescence with a few small flowers. Its flowers have an ovate-trullate lip, almost twice as long as wide, with a flattened bifold callus at the base, and a strongly and bluntly beaked anther.
Liparis trulliformis is somewhat similar to L. bathiei but different by the longer pseudobulbs, prominent plicate leaves, trowel-shaped lip which is almost twice as long as wide. It is also close to L. bosseri described above but the lip of L. bosseri is obovate (vs ovatetrullate), lacks a distinct callus at the base of the lip and has an anther with a spathulate beak vs a sharp beak. DISTRIBUTION HABITAT. Epiphyte in forest. Altitude: c. 1800 m. CONSERVATION STATUS. Category CR: the extent of occurrence (EOO) of Liparis trulliformis cannot be estimated since it is only known from one subpopulation whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). Liparis trulliformis is only known from one subpopulation representing one location (sensu IUCN), this rare species has thus been preliminarily assessed as CR using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slashand-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. January. ETYMOLOGY. Referring to the trowel shape of the lip. NOTES. Known from the type specimen only, collected by Perrier in 1923 on Mt Tsaratanana close to the type locality of Liparis zaratananae (the collecting numbers are consecutive). Only four plants grew in the colony and two were collected and sent to Schlechter in 1923 who described it a year later together with several other Liparis. These might be juvenile plants of another species, but as it comes from a fairly underexplored area it has been decided to recognise the species until more information becomes available. ILLUSTRATION: Fig. 66.
The habit of the plant is somewhat similar to other bifoliate Liparis, including L. bernieri, L. listeroides and L. longicaulis, but its pseudobulbs are generally shorter and the leaves thicker. The lip shows some similarity to that of L. ochracea, L. longicaulis and L. puncticulata but the habit, flower, lip calli and falcate column wings are all very different, as explained in Hermans & Cribb (2014: 4). DISTRIBUTION 4.5; C & D lateral sepal × 6; E dorsal sepal × 6; F, G petal × 6; H lip, side view × 6; J lip from above × 6; K lip basal callus × 9; L lip and column from side × 6; M anther cap, side view × 21; N anther cap from above × 21; P anther cap underneath × 21; Q column, front view × 9; R column and callus, side view × 9; S pollinia × 21. From Hermans 4187 (K). DRAWN BY OLIVER WHALLEY. plants collected by Warpur, an inflorescence with one flower only, lacking its lip from a cultivated plant from the Warpur collection, plus an unsigned watercolour from a cultivated plant from the Warpur collection showing the plant and a close-up of a flower (Fig. 70). This combination of material clearly shows that L. clareae var. angustifolia, described by Perrier in 1936 and known only from the type specimen, is also the same taxon and quite distinct from L. clareae.
Plants were collected by Warpur in Madagascar in June 1900 and herbarium material of plants without flower arrived at Kew in December; the herbarium sheet has a label by Warpur stating that the specimen came from Tananbe (or Tamambe) and grew in soil on dry slopes in forest. One can assume that at the same time living plants from the same origin also came to Kew where they were cultivated for several years. During this time Warpur supplied great quantities of orchids from Madagascar to European orchid growers, either directly or through auctions at Protheroe & Morris. Some of these were also grown at Kew. It is not entirely certain where the plants were collected: Tananbe could be Tanambe near Lake Aloatra in Toamasina province but there are several other possibilities, the name meaning 'big village' in Malagasy.
The plants flowered at Kew on several occasions, a plant and flower were painted in 1906 (in the style of Matilda Smith, Fig. 70) and a flower without a lip is also preserved on a herbarium sheet in 1907, both are annotated with the Kew accession number 437. Rolfe finally described the new species in 1908. The herbarium sheet and watercolour are both annotated by Rolfe as being 'close to L. parva'. Rolfe, in his description, just referred to 'Warpur' as the type specimen but mentioned that the plant was 'introduced about 7 years ago'. The K herbarium sheet Warpur from 'Tananabe' 10 June 1900 is chosen here as the lectotype with the single imperfect flower Warpur 437 (K0002452145) being an lectoparatype plus the watercolour of Warpur 437 of September 1906 being an additional lectoparatype. Additional material without a number, dated October 1899, is also at K. Liparis cardiophylla H.Perrier var. angustifolia H.Perrier was described by Perrier in 1936 but lacked a Latin diagnosis, the name was validated in 2007 Hermans et al. (2007: 289). ILLUSTRATIONS. Figs 70, 71 and 72;Cribb & Hermans (2009: 147).
Medium-sized terrestrial plant up to 17 cm high, rhizome very short, pseudobulbs retain leaves for two seasons, roots flexuose, more or less pilose.
Liparis zaratananae is somewhat similar in habit and inflorescence to L. jumelleana but it is well characterised by the more ovate leaves with a more sharply defined petiole, larger flowers, smaller petals and the lip oval vs transversally elliptic and the spathulate vs pointed beak of the anther. There are similarities with L. sambiranoensis but the plant is generally a little smaller, the leaf more petiolate, two years growth remain whilst L. sambiranoensis emerges from a repent pseudobulb, the bracts are smaller (not cordate), the lip a very different in shape (shortly auriculate vs distinctly auriculate) and much shorter dentate at the front and the anther has a spathulate beak vs triangular. DISTRIBUTION. Endemic to Madagascar: Antsiranana province, Mt Tsaratanana only (Map 50). Fig. 71. Liparis warpuri. A habit × 1; B dorsal sepal × 6; C lateral sepals × 6; D petal × 6; E lip × 6, F column and ovary, side view × 9; G column, front view × 9; H T/S through lip × 6. Based on H. Perrier 17018 (P). Unfinished drawing by Oliver Whalley from sketches by J. Hermans. The lip callus shown in E, F and H is based on dried specimens and does not show the more swollen bi-or tri-lobed callus.
tion whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). Liparis zaratananae is only known from one subpopulation representing one location (sensu IUCN), this rare species has thus been preliminarily assessed as CR using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slashand-burn farming), resulting in habitat reduction and habitat quality reduction. At Mt Tsaratanana the habitat is particularly affected by clearance for cultivation of illegal drugs. FLOWERING TIME. Perrier 15743 P00095528 has October, P00095529 has December and Perrier 16484 has April. ETYMOLOGY. Refers to Mt Tsaratanana (Zaratanana), the type locality in Northern Madagascar. Perrier (1939: 269) changed the epithet to tsaratananae believing Schlechter to have made a typographic error, but there is no basis for this as Schlechter used the German version of spelling. NOTES. Liparis zaratananae is often confused in herbaria with other species. It was first collected in October 1922 by Perrier from a small colony of six plants on Mt Tsaratanana and sent to Schlechter for identification in May 1923. He returned it to the Paris herbarium in October of the same year and published the description a year later. There are two sheets from the same locality, one has a collecting date of October (X) and the other December, this could be an error but could equally be a collection from a later date. Perrier made another collection in 1924, from a different area of Mt Tsaratanana of a somewhat different form with a wider lip and more linear anther beak. No other reliable records exist of the species. ILLUSTRATION. Perrier (1939: 271). The photographs in Cribb & Hermans (2009: 149) labelled Liparis zaratananae are of L. chantaliae. Plant erect, very small, epiphyte, 6 cm, rhizome very short, roots rounded, filiform, glabrous. Pseudobulbs oval-conical with 2 leaves, 4 -7 mm, covered in degrading sheaths. Leaves ovate, 5.5 -6 × 2.5 -2.8 cm.
It is similar in its inflorescence and flowers to Liparis dryadum but it has broader and larger leaves. It also resembles in size and habit L. densa, L. lutea, L. nephrocardia and L. bathiei but, without flowers, it is not possible to be certain from the type herbarium material if it differs sufficiently. The type plant at BM has one growth and no flowers, its habit matches several species but it is not possible to determine which. Deans Cowan's watercolour does not allow identification with any certainty and we, therefore, consider this name ambiguous (nomen dubium designated here) due to the lack of reliable type material. It shares many characteristics with L. bathiei which may be the same species but there is not enough reliable material to support this. Terrestrial or rarely epiphytic, lithophytic herb. Roots hairy. Stem cylindrical to pseudobulbous, fleshy, often creeping and rooting in basal part. Leaves thin textured to fleshy, usually pleated, petiole sheathing at base. Inflorescence erect, racemose, unbranched. Floral bracts setose to lanceolate, persistent. Flowers non-resupinate or resupinate, green, brown, yellow, pink or purple. Dorsal sepal spreading, free. Lateral sepals free or fused, spreading. Petals often narrower than the sepals, free, spreading. Lip erect, flat but sometimes concave at base, entire to lobed, auriculate at base or lacking auricles, apical margins entire to toothed, lacking a spur, callus absent, cushion-like or obscurely transversely ridged. Column lacking a foot; anther dorsal, attached by a slender filament, locules opening ventrally, pollinia four, waxy, lacking appendages or rarely with one or two tiny viscidia, stigma semi-circular or ovate; rostellum often obtuse or emarginate at apex (adapted from Cribb in Pridgeon et al. 2005: 471). DISTRIBUTION. A genus of about 300 species found throughout the tropics and subtropics of the Old and New World, with a few species in temperate regions of Europe, Asia and the Americas. Four species in Madagascar of which three endemic (75%), one endemic to the Seychelles, one to the Comores, one with a wider distribution on main-land Africa. None on the Mascarenes. Ridley mentioned the W African Microstylis stelidostachya Rchb.f. (Reichenbach f. 1881: 118) (now Orestias stelidostachya (Rchb.f.) Summerh. (Summerhayes 1937: 460) from the Comoros in error. TAXONOMIC NOTES. Most authors until recently have followed the treatments of Schlechter (1911a) and Seidenfaden (1978) in accepting a broadly defined Malaxis. However, in this sense, Malaxis is undoubtedly polyphyletic. More species need to be included in ongoing DNA studies to resolve the problems.
The systematic history of Malaxis is complex. Ridley (1888) produced the first infrageneric treatment of Microstylis (subsequently subsumed in Malaxis), recognising eight sections, six of which were confined to the New World, one to Africa, and two to Asia. Smith (1909 and Schlechter (1911a) recognised non infrageneric groupings in Microstylis. Seidenfaden (1978) included Microstylis in Malaxis and discussed previous infrageneric treatments, transferring five section established in Microstylis to Malaxis.
More recently, Szlachetko (1995), Szlachetko & Margonska (2002) and  recognised a number of new genera in subtribe Malaxideae. Most of them were included in Malaxis by previous authors. Szlachetko & Olsewski (2001) accepted the segregate genera Lisowskia and Kornasia in their account of the orchids of Cameroon.
Data from DNA sequence analysis shows that Malaxis s.l. is polyphyletic (Cameron 2005) but not along the lines defined by Szlachetko and his co-workers. ETYMOLOGY. From the Greek malaxis (malasso), to soften, in allusion to the soft-textured leaves.
Lip very shortly and narrowly winged at the base below the column then abruptly broadened into a subglobose to oval blade rounded at the anterior margin, with 2 longitudinal crescent-shaped swellings along a central vein, 1.9 -2.2 × 2.1 -2.7 mm. Column very short, with characteristic large rounded wings at the apex, 0.9 -1.2 × 0.4 -0.5 mm. Anther very small, broadly obovate, with two distinct chambers, c. 0.3 × 0.35 mm. Pollinia ovoid c. 2 mm RECOGNITION. A small plant with three leaves, small flowers with an obtuse dorsal sepal, widened at the base, a lip slightly wider than long, with a very narrow base not surrounding the column, two crescentshaped swellings on the blade along a central vein, and a very short column with large rounded wings at the apex.
The broad lip is similar to that of Malaxis cardiophylla but plant and leaf are different with the base of M. cardiophylla being more cordate than this species, the leaves of M. cardiophylla also have a distinct elongate petiole. The lip is broader in M. cardiophylla which also lacks the crescent-shaped swellings. The column has rounded apical wings (vs angular). The species is different from M. weberbaueriana (Kraenzl.) Summerh. (Summerhayes 1934: 208) in having leaves with a shorter petiole, flowers that are equally small, but have a lip with much smaller basal auricles that do not surround or surpass the column as in M. weberbaueriana. It differs from M. welwitschii from Angola by the less cordate ovate leaves, the entire ovate lip (vs three-lobed) and the large rounded wings (vs angular) of the column. DISTRIBUTION HABITAT. Terrestrial in humid evergreen forest on plateau. Altitude: 1000 -1500 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Malaxis atrorubra cannot be estimated since it is only known from two subpopulations whereas its area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). With only two known subpopulations representing two locations (sensu IUCN), M. atrorubra has been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. October to January. ETYMOLOGY. Refers to the dark-red to dark purple colour of the flowers. NOTES. Perrier originally described it as Microstylis atroruber and it was transferred to the genus Malaxis by Summerhayes (1953). According to the ICN (Turland et al. 2018: 60.11) the use of a hyphen in this instance is to be treated as an error and corrected: therefore atro-rubra becomes atrorubra.
Perrier's holotype is a young plant with few flowers. In his description he noted that the lip calli were verrucose which is not obvious in specimens. Margonska (2009: 91) considered this species to be conspecific with Malaxis welwitschii from Angola (as Lisowskia welwitschii) together with M. francoisii (H.Perrier) Summerh. (Summerhayes 1954: 578) also from Madagascar but the plant, lip and column of M. atrorubra are considerably different as explained above. The species was described at the same time as M. francoisii and based on a collection by Perrier and François, who would have seen the living plants. It is unlikely that Perrier would have described the same species twice in the same paper. ILLUSTRATIONS. Perrier (1939: 255).
Malaxis cardiophylla is similar to M. weberbaueriana but the plant and flowers are distinct, especially the leaves which are cordate (vs lanceolate) and a lip which is broader and does not surpass or envelop the column. A drawing by Ridley (at K) based on Humblot 437 (presumably based on the BM specimen), shows differently shaped (not cordate) leaves. Perhaps M. cardiophylla is an aberrant or local form of M. weberbaueriana but based on current evidence it is left as a separate entity. It is also similar in lip-shape to M. atrorubra but it is broadly vs narrowly winged, its habit and leaf shape are slightly different with a broadly cordate leaf vs ovate with the base oblique, and its column structure with rounded (vs angular) lobes. DISTRIBUTION. Comoro Islands, Mayotte. Perrier (1939: 257) gave 'Grande Comore, Combani' as the locality of this species but Humblot's Field Notebooks in P (1882 -1885) clearly list the "plant growing next to rocks, flowers violet, from Mayotte, Combani forest". It is a common mistake to confuse Combani forest of Grande Comore with Combani on the island of Mayotte (Hervouet & Barthelat 2014: 268) (Map 53). SPECIMENS EXAMINED. COMOROS. Mayotte, Combani forest, Humblot 1437 (437) (holotype W, isotypes BM, LD, LG, P (P00094908-10)); Mayotte, Grande Terre, Tsararano, Benara, Dec. 2002, Barthelat 1116 (K, P); Mayotte, Summit of Petit Benara, Jan. 1990, Tinguy 1134 (P). HABITAT. Near rock, in forest, terrestrial or epiphytic. Altitude: around 300 -400 m. CONSERVATION STATUS. Category CR: The extent of occurrence (EOO) of Malaxis cardiophylla cannot be estimated since it is only known from one subpopulation whereas its minimal area of occupancy (AOO) is estimated to be 4 km 2 (which falls within the limits for Critically Endangered status under the criterion B2). Malaxis cardiophylla is only known from a single subpopulation representing one location (sensu IUCN), this rare species has thus been preliminarily assessed as CR using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence, slash-and-burn farming and urban development, resulting in habitat reduction and habitat quality reduction.
Previously assessed and published in the IUCN Red data listing for Mayotte as VU D2 (Mercks 2013). FLOWERING TIME. December to January. ETYMOLOGY. A reference to its heart-shaped leaves. NOTES. Reichenbach originally described it as Microstylis cardiophylla but it was transferred to the genus Malaxis by Kuntze (1891). The species was placed in the genus Orestias by Margonska & shortly acute, c. 4 × 0.35 mm. Lip 3-lobed, the lobes obtuse, the lateral lobes auriculate the mid-lobe a little bigger, with the auricles at the base as long as the lobes and completely enveloping the column, midlobe with a rounded apex, c. 5 × 4.5 mm. Column short, wings small, obtuse. RECOGNITION. Malaxis madagascariensis is a small plant with conical-cylindrical pseudobulbs, pleated undulate leaves, and a long inflorescence with up to 30 flowers.
The flowers are small, non-resupinate, and have a distinctly three-lobed lip with the basal auricles surrounding the column.
It has some similarities with Malaxis physuroides but in M. madagascariensis the pseudobulbs are thicker and shorter, the lip is three-lobed (vs suborbicular). There are some vegetative similarities with M. seychellarum but again the three-lobed lip with entire margins vs oblong, distinctly dentate at the margin lip is very different. In many ways the species is more akin to some E Asian species, such as Crepidium punctatum (J.J.Wood) Szlach. (Szlachetko 1995: 131), a species that also has spotted leaves, a similar habit but a different flower structure with the dorsal sepal broader and the lip far less deeply three-lobed. DISTRIBUTION ETYMOLOGY. Referring to the assumed country of origin. NOTES. Klinge originally described it as Microstylis madagascariensis but Summerhayes (1954) transferred it to the genus Malaxis. : 303) placed the species under Crepidium koordersii (J.J.Sm.) Szlach. (Szlachetko 1995: 127) from Indonesia. Margonska does not substantiate this decision apart from stating that "they are without any doubt conspecific and that no species from Crepidium has ever been recorded from Madagascar and that the origin of the Microstylis madagascariensis plant cultivated at Hort. Petrop. was Madagascar and the subsequent mechanical transferring of taxa without careful checking have led to citation errors in contemporary publications relating to Madagascar". Although the plant habit and lip have some similarities there are significant differences between the two species: the leaves of Malaxis madagascariensis are considerably like sheaths, red. Rachis very densely racemose. Floral bracts erect, narrowly lanceolate, acute, up to 2.3 × 0.5 mm. Flowers not resupinate, c. 3 × 4 mm, dark red. Pedicel and ovary glabrous 2 -2.3 × 0.2 -0.4 mm. Dorsal sepal oval-lanceolate obtuse, 1.7 -2 × 0.8 -0.9 mm. Lateral sepals oval-obtuse, 1.2 -1.7 × 0.7 -0.8 mm. Petals linear, 1 -1.2 × 0.2 -0.3 mm. Lip suborbicular, obtuse with distinct obtuse basal wings folded around the column, margins of the blade sinuate, anterior margin rounded-truncate to a little indented, disk with a subreniform callus (perhaps pubescent) just below the middle, 1.6 -1.9 × 1.5 -1.7 mm. Column short, straight, with indistinct wings, c. 0.5 × 0.2 mm. Anther ovate, truncate at the back, angular-obtuse at the front c. 0.33 × 0.2 mm. Pollinia ovoid c. 2 mm. RECOGNITION. A small plant with cylindrical succulent pseudobulbs, two to three broadly apical oval leaves. Inflorescence almost as long as the plant, very denselyflowered, flowers small, red, lip suborbicular with obtuse basal wings alongside the column, anterior margin rounded-truncate with a distinct rounded callus on the disc.
Schlechter, in his description, likened it to Microstylis stelidostachya Rchb.f. and Malaxis katochilus Schltr. from E Africa (now Malaxis prorepens (Kraenzl.) Summerh. (Summerhayes 1934: 208) but the rachis is much less dense and the flowers are different especially in the shape of the lip and the rounded callus. Malaxis physuroides differs significantly from all the other Malaxis from the Madagascar region by the very dense rachis of small flowers and the distinct rounded single callus on the lip. DISTRIBUTION Nov. 2007. HABITAT. Humid evergreen forest, shaded wet rocks. Altitude: 400 -1400 m. CONSERVATION STATUS. Category EN: the extent of occurrence (EOO) of Malaxis physuroides is estimated to be 31,200 km 2 (which falls within the limits for Endangered status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 20 km 2 (which also falls within the limits for Endangered status under the criterion B2). Malaxis physuroides is only known from three subpopulations representing three locations (sensu IUCN), this species has thus been preliminarily assessed as EN using the green listing method. This species is threatened by selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming), resulting in habitat reduction and habitat quality reduction. FLOWERING TIME. November to April. ETYMOLOGY. A reference to the orchid genus Physurus Rich. ex Lindl. (now Erythrodes Blume (1825: 410)) with which it shares the habit of some of the species. NOTES. In 1913 Schlechter described it as Microstylis physuroides, based on material collected by Perrier and also illustrated the plant and flower. Schlechter, in his early collaboration with Perrier, used a different numbering system for specimens than Perrier. It was placed into the genus Malaxis by Summerhayes (1954). Szlachetko (1995) transferred it to his new genus Lisowskia but none of the morphology or genetics support this. Margonska (in Margonska et al. 2012: 454) designated as lectotype P00094916, referring to Perrier's specimen in Paris 8010 (89 of Schlechter). She designated a hypothetical specimen in Berlin that was assumed lost as an isolectotype, but there is no reason to believe that such a specimen existed or was lost. As there is only one Perrier sheet (annotated by Schlechter) that is obviously original material used by the author, there is no need to assign lectotypes. ILLUSTRATIONS: Schlechter (1913: pl.7;1932: T.51). Medium to large humicole plant, epiphyte or lithophyte 10 -33 cm tall but generally between 15 and 25 cm, rhizome very short, repent, roots filiform, flexuose more or less pilose. Pseudobulbs stem-like, narrowly cylindrical-ovoid to subcylindrical, thickened-bulbous at the base, 5 -8 cm long, 1 -2 cm in diam., covered by 2 -3 long and overlapping stem-sheaths, sometimes the older pseudobulb obliquely divergent and always leafless, with up to 2 -5 leaves but generally 3 -4, starting half-way up the stem and ending with 2 -3 larger ones towards the apex. Leaves erectly spreading, obliquely elliptic to ovate-elliptic, apiculate to obtuse, base rounded-cuneate and then attenuate into a 1.5 -3 cm petiole, overall 7 -20 × 2.5 -7 cm, plicate, margins a little undulate to crenulated, thin, purple becoming green or green. Inflorescence erect, up to 30 cm, 2 -3 mm in diam., green to purple, densely racemose with up to 50 flowers. Peduncle costate, with 2 -3 bract-like peduncle sheaths, up to 4 × 2 mm. Rachis around ¾ of the inflorescence. Floral bracts arched, descending, linear-lanceolate-acuminate, 3 -8 × 1.5 -2 mm, green to purple-brown. Flowers not resupinate, very small to small, opening in succession, overall c. 5 × 6 mm, variable in colour from pale greenish-yellow, lemon-yellow, purplish green to purple, fading to yellow-orange, there seem to be distinct colour forms in different locations. Pedicel and ovary linear, slightly arched, 3.2 -5 × 0.3 -0.8 mm, green, often tinted purple. Dorsal sepal lanceolate, obtuse, 2.3 -3.9 × 0.4 × 1.1 mm. Lateral sepals linear-oblong, obtuse, subfalcate, 2.3 -3.2 × 0.9 -1.6 mm. Petals linear, obtuse, arched and crossing below the column, 2.9 -3.8 × 0.3 -0.6 mm. Lip transversally oblong to flabellate, shortly winged at the base, with a more or less distinct basal callus, distinctly irregularly dentate at the anterior margin, 1.8 -2.1 × 2.5 -3.3 mm. Column slightly incurved, wings small, distinctly rounded, apex triangular, 1.2 -1.6