- 8 — 



Moreover the habit is that of a Gastrochilus and the violet lip with 

 radiating white lines shows more relation to that genus than to Curcuma. 

 At all means it is to be excluded from Curcuma. 



5 C. alismatifolia Gagnep. ! 1908, 57 (v.s. Herb, mus Paris 343) has 

 at first look a rather striking resemblance to a true Curcuma because of 

 the long pink coma lobes. On the other hand it resembles C. parvifolia 

 by the violet colour of the labellum and the shape of the bracts, both rather 

 different from a true Curcuma. The central furrow in the lip points also 

 more to a Gastrochilus then to a Curcuma but it reminds C. Roscoeana. 

 The stamen differs as well horn thai oi C. parvi/lora as irom a true Curcuma. 

 The narrow parallel thecae of the rather long crested anther are attenuate 

 at their base into a kind of spurs and the connexion with the filament is 

 at the backside near the base, probably it is nutant. The pointed thecae 

 remind some species of Gastrochilus, but also Curcuma meraukensis. 



6. C. sparganifolia Gagnep.! 1908, 59 (v.s. Herb, mus Paris, 30). Here 

 the bracts of the spike are quite free one from another and herein they 

 differ essentially from those of Curcuma. The anther with the shortly pointed 

 thecae, is evidently terminal. The staminodes are free from the filament. 

 The labellum is entire, orange-coloured in the centre. 



7. C. gracillima Gagnep.! 1908, 59 (v.s. in Herb, mus Paris, leg. Pierre 

 and Harmand.) Here the bracts are all alike, erect, with extant subacute 

 tips. The anther is not spurred, shortly crested, wether it is terminal or 

 versatile is unknown. 



The lip is bibbed. Colour unknown. 



The three last named species have in common a very curious peculiarity : 

 The stylodes at the bottom of the flower are failing; at least they are not 

 to be observed in the dried flower, as was stated by Gagnepain and as 1 

 could persuade myself by examining flowerbuds of th-e above cited materials. 

 This peculiarity was never observed in any species of the Order, except 

 by Gagnepain in Kaemferia cuneata, Gagn. (1905, 546) a species very near 

 to K. elegans. I, however, am almost sure' that if living materials were 

 examined, these organs, either very small or connate with other organs, 

 would be found. Now I think it a very important circumstance, communicated 

 by Wallich that in the living flower of C. /?oscoéana the stylodes or nectaries 

 are exceedingly small and only are to be traced by their yellow colour. 

 Undoubtedly this shall prove to be the case in the above named species, 

 and it evidently indicates a certain relation between the three here named 

 species and C. Roscoeana. 



8. C. sylvestris, Ridl (1893, 73) (v.s. Herb, mus Paris ex herb. Pierre!) 

 Slender creeping rhizome. Scape accompanied by a solitary leaf. Flower 

 resembling that of a Gastrochilus with terminal anther, with a recurved 

 violet crest, and an emarginate lip with a yellow central spot and violet 



