- 26 — 



the stamens thicken in their central part and contract in a corkscrew fashion 

 while dying and so the anther gets difformed. Again if the buds examined 

 be too young the proportions of the anther are abnormal. So e.g. the anther 

 described by Gagnepain in C. longa (1908, 63) = C. domes^/cû, Val. showed 

 the connective almost quite loosened from the thecae which were attached 

 to the filament near their top. This is quite contrary to what is seen in 

 living flowers of that species. 



3e Pistillum. The ovary is hairy in all species of Curcuma known to 

 me (according to Schumann this is not the case in C. /^ucor/z/za, Roxb.) and 

 so is the base of the calyx too. The ovules are apparently always well- 

 developed; on the top of the ovary are to be found the two cylindrical 

 nectaries or "stylodes", varying in lenght between 4—8 mm. and secreting 

 a large quantity of nectar, which remains stored up in the corolla-tube. 



In ancient descriptions (Hooker, Bentley and Trimen) these are often 

 erroneously called staminodia. 



The style is filiform, glabrous, runs along the back of the faux and 

 filament, passes between the thecae and terminates in the stigma. This is 

 a cup-shaped two-lipped organ with a broad transversal chink, the ciliated 

 lower lip of which is somewhat protruded. 



For the description of fruit and seed compare C. aurantiaca. 



Fertilization. 



The flower of Curcuma, as well of Eucurcuma as of Paracurcuma, is 

 quite fitted for pollination by insects, as also is known from the other 

 Zingiberaceae, and especially for cross-fertilization. In one ?,ptc\t?,(C.soloensis) 

 the loculi of the anthers- contain no, or a very little quantity of, pollen 

 connected by a narrow strip of tissue and these are for the rest filled with 

 mucilage. Nevertheless 1 once found pollen on the stigma of a flower of this 

 species. It must have been brought from the flower of an allied species, of 

 which many specimens grew in the neighbourhood. In our cultivated species 

 however the pollination happened very rarely; they flowered in the middle 

 of the rainy season and insect visits were very rare. 



Only once during the many months in which 1 observed the flowers I 

 saw a green bee (probably an Anthophora) enter into two flowers of 

 C. euchroma, and a Xylocopa visiting flowers of Cai/ran/Zaco. Where pollination 

 occurs an abundant crossing may be expected between related species for 

 the form and construction of the flower is nearly equal in many species of 

 Eucurcuma. I believe, however, that fertilization is also very rare in wild growing 

 plants, e.g. in the very common and abundantly flowering C. purpurascens, wild 

 growing species belonging to the Mesaniha. For fruits never were found in any 

 of the spikes of central floweringspeciescollected in theteakforests. Only two 

 species: Curcuma Mangga, var. sylvestris (an Exanthous species) and Curcuma 

 aurantiaca v. Zyp (a Paracurcuma) I received abundantly fructiferous spikes, 



