NOTES RELATING TO THE FLORA OF BHUTAN: XLIV. TAXONOMIC NOTES, NEW TAXA AND ADDITIONS TO THE ORCHIDACEAE OF BHUTAN AND SIKKIM ( INDIA)

The terrestrial orchid genus Bhutanthera Renz (subfamily Orchidoideae , tribe Orchideae ), comprising ﬁve species, three newly described here, is established. A new species, Herminium pygmaeum Renz, and three new varieties, Bulbophyllum cauliﬂorum Hook.f. var. sikkimense N. Pearce & P. J. Cribb, Cephalanthera erecta (Thunb.) Bl. var. oblanceolata N. Pearce & P. J. Cribb, and Gymnadenia orchidis Lindl. var. pantlingii Renz, are here newly described. Five new combinations and one new name are proposed. Four new sections of Bulbophyllum are also established and their relationships discussed.

Etymology.From Bhutan, where most of the species are found.
The genus Bhutanthera comprises five terrestrial species confined to the alpine zone of the eastern Himalayas.It falls between Habenaria and Platanthera but is distinguished by having globose tubers, a 3-lobed lip and conjoined cushion-like stigmas.
The three new species of Bhutanthera described here, along with Habenaria albomarginata King & Pantling and H. alpina Hand.-Mazz., represent a difficult taxonomic group.Habenaria albomarginata, for example, has also been placed in Platanthera and Peristylus.Renz (personal communication) suggested that these five taxa might represent a sufficiently distinct taxonomic group to justify the establishment of a new genus for which he proposed the name Bhutanthera.He considered that it was related to Platanthera but differed in having a 3-lobed lip.This feature is not found in other Platanthera species from the region but is found in the genus elsewhere.
Some botanists (Inoue & Lin, 1980;Inoue, 1988) consider that the tuber shape is of particular taxonomic significance in the subtribe Orchidinae.On this basis alone it is difficult to justify the inclusion of this group in Platanthera.Luo (personal communication) considers that their distinctly globose tubers place them closer to the Habenaria-Peristylus complex.When King & Pantling described H. albomarginata they placed it in subgroup Peristylus (now considered a distinct genus).The globose tubers and 3-lobed lip of these five species place them nearer to Habenaria and Peristylus but they cannot be comfortably accommodated in either because of their conjoined cushion-like stigmas.
A more detailed investigation of the entire complex is undoubtedly needed before the complex can be satisfactorily resolved.Renz considered that the establishment of a new genus to accommodate this group was the most satisfactory solution given the present state of knowledge of the subtribe, and we concur.
Distribution and ecology.India (Sikkim).On alpine meadows and yak-grazed slopes, 3720-4270m.Flowering between July and September.Bhutanthera albomarginata, selected as the type for the genus, is well illustrated on plate 425 by King & Pantling (1898).They originally described it as Habenaria albomarginata, commenting that 'the centrifugal inflorescence is a very unusual feature in Habenaria'.It was transferred by Kra ¨nzlin (1898) to Platanthera with the comment 'Erinnert habituell an ein etwas gestauchtes massiv gebautes Exemplar von Platanthera viridis Lindl.' (Platanthera viridis is now treated as Coeloglossum viride (L.) Hartman).Finally, Banerji & Pradhan (1984) transferred it to Peristylus but without any explanation.
Distribution and ecology.Bhutan.In rock crevices, on cliffs, 4100m.Flowering in August.
Bhutanthera albosanguinea is related to B. albomarginata but differs in having a 2-flowered inflorescence, white sepals, red petals and a red lip.The type sheet consists of a single, slightly damaged plant.
Distribution and ecology.India (Darjeeling and Sikkim).In tropical valleys, 660-2000m.Flowering June to August.Examination of the type material of B. cauliflorum shows that there is variability in the number of veins on the petals.Griffith 5139 has distinctly lanceolate petals with three veins while Griffith 5165 and Hooker 36B both have petals with a single vein.All of the material from Sikkim has flowers with petals that have a single vein.We believe that this represents a local variation and propose varietal status for these plants.We propose to lectotypify B. cauliflorum with Griffith 5139, while Griffith 5165 and Hooker 36B are both referable to B. cauliflorum var.sikkimense.
King & Pantling (1898) identified Pantling 230, from Sikkim, as B. protractum Hook.f. but noted that his plants had petals with a single vein.We have examined this material and believe that it is referable to B. cauliflorum var.sikkimense and that B. protractum does not occur in our area.
Bulbophyllum cauliflorum is closely related to B. collettii King & Pantling from Assam.King & Pantling (1897) stated that the latter grows in dense masses, is a smaller plant with different sepals and petals and that there is a difference of two months in the flowering time.Pantling collected a plant from Shillong that was flowering in June.The type material of B. collettii (Assam, G. Rita & H. Collett) is at CAL and we have been unable to examine it.Further collections will have to be examined before its status can be clarified.
The floral structure of B. cauliflorum var.sikkimense resembles that of B. laxiflorum (Bl.)Lindl.but that species has clustered, rather than well-spaced, pseudobulbs.
Material of this variety is identical in habit to that of the typical variety but the lip is simple, unadorned and lacks a spur.We believe that it is a peloric variant of C. erecta.We have examined several collections and they uniformly have such flowers.The typical variety has yet to be collected in Bhutan.Lindl., Gen. Sp. Orchid. Pl.: 278 (1835).Type: India, Kumaon, Wallich 7039B ( lecto.K-LINDL!, iso.K-W! K!, selected here).Fig. 7.

Habenaria diphylla
H. clarkei Kra ¨nzlin in Bot.Jahrb.Syst.16: 215 (1893).Type: India, Sikkim, Hooker Herb.Ind. Or. 42 (holo.B †, iso.K!, LE?).Renz (personal communication) considered H. josephi to be conspecific with H. diphylla.However, it differs from H. diphylla in that the stem is ebracteate while in H. diphylla there are many bracts above the leaves.It differs from H. aitchisonii, where it was placed as a variety by Joseph Hooker (1890), in having coiled tips to the lateral lobes of the lip.
actually refers to part of the type collection of both H. josephi and H. josephi var.aitchisonii.The type material referred to by Kra ¨nzlin has probably been destroyed.The description closely resembles H. diphylla but the petals are described as 'bipartitis' which is not the case with H. diphylla.We are sure that the specimen referred to by Kra ¨nzlin is part of the type material referred to above and are therefore confident that it can be reduced to synonymy here.
There has been considerable debate as to the identity of this taxon since Rolfe (1905) distinguished T .venosa from plants previously grouped under T .alba (Lindl.)Rchb.f.Seidenfaden (1986) reviewed the position but it is worth pointing out that the Wallich plate of Phaius albus (t.198, 1831) has a lip with a yellow centre and pink branching veins.The only consistent difference between T .bracteata and T .alba seems to be the absence of yellow on the lip in the flowers of the former.The keels have been described as obscure but a photograph at Kew of T .bracteata clearly shows long, fimbriate keels on the lip.It has been impossible to identify any authentic specimens of var.bracteata from Bhutan and Sikkim but Pradhan (1979) reported it from Sikkim and Darjeeling as T .venosa Rolfe.The original collection from the Garrow [Garo] Hills has probably been lost and Seidenfaden (1986) proposed the above neotypification.
Because the differences in vegetative and floral morphology between T .alba and T .bracteata are so slight we here propose that the plants lacking the yellow patch on the lip be treated as a variety of T .alba.Clearly further collections are required to test this solution.Distribution and ecology.India (Sikkim).1200m.Flowering August.

Goodyera clavata
King & Pantling (1898) described this species giving it the name G. grandis.However, they had apparently overlooked the earlier publication of the epithet by Blume for Neottia grandis when describing a plant from Bantam in Java in 1825.Later, in 1858, Blume transferred it to Goodyera as G. grandis in his Flora Javae et Insularum adjacentium.King & Pantling mentioned that the plant they described differed from all other Goodyera species by its characteristic clavate callosity on the anterior surface of the column below the stigma.We have examined the type specimen (Pantling 460) and material of G. grandis Bl. from Java and are confident that they are different species.The Himalayan plant has a lip lacking a claw, its apex is small, reflexed and adpressed to the sac which has setose swellings laterally, and the column has the characteristic clavate swelling.The Javanese plants have a lip with a central claw, a much longer recurved apex that is not adpressed to the sac, that has setose swellings mainly at the junction between the sac and the mesochile and a column lacking a clavate swelling.King & Pantling (1896) commented that their new species, Habenaria cumminsiana, was related to H. pachycaulon Hook.f. and belonged within section Hologlossa Hook.f.This section is characterized by plants with entire petals and a simple, entire, linear lip.Many of the species within this section have already been transferred to Platanthera, including H. pachycaulon Hook.f.(as P. pachycaulon (Hook.f.) Soo ´).The elongate rhizomatous tuber and the simple lip of H. cumminsiana, both characteristic of Platanthera species from the Himalayas, confirm that it belongs in Platanthera.
The taxonomy of this species has been the source of confusion for some time.Bru ¨hl (1926) based the description of his monotypic genus, Cleisocentron, on that of King & Pantling (1898).He selected Saccolabium trichromum as the type and reduced S. pallens to synonymy.Garay (in Seidenfaden, 1992) considered that these might be two distinct species based upon the difference in the appearance of the columnfoot.Seidenfaden (1992) mentioned that the plate in Xenia Orchidacea (2: 119, t.139 (1874)) of S. trichromum shows no clear column-foot while the plate in King & Pantling (1898) showed a distinct column-foot.The illustration of Cathcart associated with the description of S. pallens clearly shows a column-foot and in all the Sikkim material we have seen has this feature.We believe that the plate in Xenia Orchidacea probably does indicate a decurrent column-foot by the constriction of the column into two equal parts.We have carefully examined these taxa and believe them to be conspecific.Garay (personal communication) agrees with this position.
There is an additional problem concerning the correct name for this species.The publication of S. pallens by Lindley was issued on 20 August 1858, while the reference by Reichenbach relating to S. trichromum, was published in February 1859.Therefore, the name S. pallens takes precedence over S. trichomum and the new combination based on the former is made above.

Bulbophyllum Thou.
We have chosen to accept Bulbophyllum as a broadly defined genus in the Flora of Bhutan treatment, whilst realizing that current work on the phylogeny of tribe Bulbophyllinae may lead to its narrower circumscription as advocated, for example, by Garay et al. (1994).Four new sections of Bulbophyllum are needed to accommodate Himalayan species of the genus that occur in the Flora of Bhutan region.These are as follows: Bulbophyllum section Biseta J. J. Vermeulen, sect.nov.Affinis Bulbophyllo sectioni Racemosis sed folio unico, inflorescentia basali, alis lateralibus duabus filiformibus elongatis ex ovario orientibus, petalis spathulatis et labello integro differt.Type: B. bisetum Lindl., 1842.
Allied to Bulbophyllum sect.Racemosa Benth.& Hook.f. but differs in having unifoliate pseudobulbs, two characteristic filiform long lateral projections ( jugae) that arise laterally from the ovary, spathulate petals and an entire lip.