flower is much wider and the dorsal petal is much shorter than the lip 

 and as wide. The anthers are much resembling in both species. In both 

 the top is emarginate the thecae long linear and densely villous splitting 

 along the whole front, but in H. lycostoma the connective continues a little 

 beyond each of the thecae, forming a short bilobed or deeply emarginate 

 a little recurved appendix, while in H. minor there is only a thickened 

 emarginate recurved edge. 



The calyx is red in H. lycostoma, snowwhite, beautifully gauzy reti- 

 culate in H. minor. The flowerbracts are oblong acute, thin glabrous as 

 long as the flower, white in H. minor, partly red in H. lycostoma. The 

 bracteoles are very small and the innermost ones 'almost obsolete. 



The stylodes are peculiar in both species, they are short (2 mm.) and 

 connected to a hollow cylinder around the style, composed of numerous 

 (5—8) oblong conoid corpuscules free at the top and of rather inequal 

 size; but their number does not surpass 5 in H. lycostoma and arises to 8 

 in H. minor. SCHUMANN calls the ,, petals" white; in all our specimens 

 however and also in the original spec, of LAUTERBACH 2542 !, quoted 

 by K. SCH , the petals and the whole flower we're red only the anther was 

 white, and an obtuse crest of the anther as mentioned by K. SCHUMANN 

 does not exist either here or in any species of this group. 



This species, which is very common and largely dispersed in the pa- 

 puan archipel shows more or less important varieties. So the mountainous 

 form collected by Mr. De Kock, has larger more squarrose spikes, larger 

 and wider flowers shorter petals, etc. than the original form from lower 

 places. There is also variation in the reticulation of the sheaths. 



Horntedtia alliacea VAL. in Ic. bog, (1912) t. 350.— Donacodes 

 alliacea T. et B.! Cat. H. bog. 1866, sine descr.-? H. conica RlDL. (1899), 124. 



Distribution. 



Java: Batavia, Buitenzorg, sea level to 100 M. Cult, in pagia. (nom. 

 ind. Pining bawang). Prinseneiland. (leg. NATHANSON in Herb. bog. spont. 

 Cult, in Hort. Bog. (Nursery). 



Sumatra: Lampong, Seputih-Tulong-Bawang,-Kuta-Bumi. (GUSDORFF, 

 Febr. 1914, No. 177 ,, Between other plants of the order socially in swampy 

 jungle along rivulets"; vern. name: Ketimbang ketanim). 



? Riouw: Pulu Bintang. (leg. TEYSMANN !). Determination uncertain. 



? Malacca. (RIDLEY, H. conica). 



? Philippines. (ELMER 10246 in Herb, bog, labelled M. conica RlDL.) The 

 fragmentary specimen cannot be identified, but the silky covering of the 

 bracts reminds H. alliacea and H. deliana Val. The villosity ol the leaf 

 is more like the latter. 



To my full description (Ic. 350) 1 have to add still a few details. The 

 bracts here are ovate, the inner sterile bracts which are very numerous are 

 oblong acute. All are silvery sericeous over the whole surface, only red near 



