202 Materials for a Flora of the Malayan Peninsula. 



being either quite rudimentary or absent : these male flowers are about 

 half as long again as the hermaphrodites. The genus Bapania was 

 founded by Korthals to receive the Sumatran plant which that author 

 called D. racemosa. Of this, there is an authentic specimen in the 

 Calcutta Herbarium ; bat unfortunately it has neither flowers nor fruit. 

 Amongst Forbes' plants collected in Sumatra, ISTo. 1217 agrees so per- 

 fectly with Korthal's D. racemosa in foliage, as to leave no doubt that 

 it belongs to the same species ; aud on Forbes' specimens there are ripe 

 fruits. These ripe fruits, as well as its leaves, show D. racemosa, to be quite 

 distinct from D. scandens, Stapf, the leaves being thinner in texture, 

 and the fruits longer than those of D. scandens. The seeds of Forbes' 

 specimens do not, however, show the curious aril (laciniate and almost 

 2-lipped) which Korthals describes : and, as the seeds of D. scandens 

 show no trace of an aril, I agree with Dr. Stapf that Korthals 

 probably described the existence of an aril as the result of some 

 confusion. The majority of the ovaries dissected by me have only a 

 single ovule : in one or two cases, however, a second (as figured by 

 Dr. Stapf) has been found. In no case, however, have I found two seeds 

 in a loculus of the fruit. There is no doubt that, as Dr. Stapf points 

 out, the genus Gonnaroipsis comes very near Bapania; and it may 

 become desirable, as both genera became better known, to reduce 

 Connaropsis (which dates only from J 862) to Bapania which was 

 published in 18-54. In the meantime the fruit of Connaropsis is not 

 properly known, and there is no evidence that it is dehiscent. The 

 fruit of C. monophylla is very like that of Bapania scandens., but no 

 specimens that I have seen show any sign of dehiscence ; while that of 

 C. macrophylla is but little angled externally and the appearance of all 

 the specimens I have seen is suggestive of indehiscence. Moreover, all 

 the species of Da^ajtia appear to be scandent ; while all those of Conna- 

 ropsis are shrubby or arboreous. Concerning the structure of the seeds 

 I can say nothing, not having met with good seeds of either. The two 

 genera remain at present separated chiefly by these two points of 

 diiference, viz., dehiscence of the fruit and habit. Whether others may 

 be found remains to be seen when better materials shall be obtained. 



Beccari's specimen (P. S. 900) appears to belong to a .species 

 slightly different from either B. scandens or D. racemosa. And, as I 

 undei-stand from Dr. Stapf, the same Collector's No. 2951, from Borneo 

 (which I have not myself examined), belongs to still another species. 



5. Impatiens, Linn. 

 Herbs, rarely shrubby at the base. Leaves opposite or alternate, 

 in some whorled, in others all radical, simple, exstipulate, or with 

 444 



