(iomjuonly 4 seeds, globose, irregularly compressed, narrowed rather abruptly 

 at the base. Arillus multi-fid, with rather broad acute linear segments, some 

 of which are longer than the seed. Fruit 12 X 9, seed 3 mm long. The fruit 

 is grey, with pale atropurpureus spots. The seed avellaneous. 



The dehiscence of the capsule is threevalved, septifrage, the valves get 

 loose and revolve backward; they are of a spongy consistence ^and soaked 

 with water at the moment of dehiscence. 



As well hereby as in the shape of the seeds there is much resem- 

 blance with Gastrochilus javanum. 



Camptandra Ridl. 



Ridley in As. S. Str. br. 1899, 103; 1907, 10; K. Sch 1904, 62. — /Taem- 

 pferia spec Baker 1890, 223, — Kaempferia p.p. K. Sch. 1904, 74. — Pjr^o- 

 phyllum. Gagn 1901, p. Lxxvii. 



This genus was based by Ridley principally on 3 characteris, viz: 

 Tlie involucrum of the in florescence. The recurvate subcalcarate, versatile 

 anther and tlie fusiform falcaie innumerous seeds. For the rest it resembles 

 closely Kaempferia, especially K. elegans and K. secunda. \n ihtïnsi named 

 the involucrum has almost the same shape, only the margins are not 

 connate at the base, but convolute. K. Schumann has reduced to this 

 genus two species from China described by Gagnepain, though here the 

 anther cells are parallel not recurved and not spurred, as appears from 

 the exact description of Gagneapain. 1 examined an original specimen of 

 C. yunaanensis, Gagn from the Mus. de Paris 2721, and found the descrip- 

 tion of the anthers quite correct, only there must be added that the anther is 

 versatile because of the insertion of the thin filament to the anther almost 

 as in Curcuwa. So this species really belongs in the genus, only the 

 expression "thecae semilunari-curvatae basi calcaratae^' should be struck 

 out froin the diagnose. This species relates to Camptandra parvula as Cur- 

 cuma aurantiaca relates to Eu-curcuma species. 



in the Herbarium bogor. 1 found a species of this genus collected by 

 Teysmann in Borneo, fruiting but without flowers, which seems to accord 

 rather well with C. parvula Var. anguslifolia, Ridl. Kaempferia gracillima 

 K. Schumann is a nearly related spec, only liffering by the presence of an abbre- 

 viated blade at the end of the spathe. It might be only a variety of Cparru/o. 

 There is a third species from Borneo in the Herb. Bog. collected by Nieuwenhuis 

 (sterile) with erect, rather thick, stems, coated in the leafless lower part with large 

 embracing bladeless lacerate sheaths and with subsessile, moderate, widely 

 elliptic lanceolate leaves 3 — 5 approximate, embracing the nodes with their 

 large sheaths. 



