190 THE JOURNAL OF BOTANY 



CyTINACEJ5. 



Besides half a dozen species of Hydnora, this natural order is 

 represented in Tropical Africa only by Pilostyles cethiopica Welw., 

 a plant hitherto supposed restricted to Angola. The extension of 

 its range to Rhodesia is therefore a matter of considerable interest, 

 and there is no doubt of this, as specimens, both male and female, 

 of a Cytinus sent from Victoria in that province by Mr. G. H. F. 

 Monro (nos. 457, 962, in Herb. Mus. Brit.) are certainly referable 

 to P. cethiopica. 



The host of this parasite in Angola is Berlinia paniculata 

 Benth. Of Mr. Monro's two specimens, the host of the female 

 has lost its leaves, and being without flowers or fruit is therefore 

 indeterminable. But the host of the male, also flowerless though 

 well provided with leaves, as Mr. Edmund Baker pointed out to 

 me, appears to be a Brachystcgia distinct from all at the British 

 Museum, and probably belongs to an undescribed species. 



2. Remarks on the Genus NEPENTHANDBA S. Moore. 



In the Journal of Botany for 1905 (p. 149, tab. 471) is de- 

 scribed and figured as a new generic type under the above name a 

 remarkable plant found by the late Colonel Beddome in Tenasserim. 

 Its chief peculiarity lies in this, that with the male flowers of Tri- 

 gonostemon, the calyx of the female flowers is largely accrescent, in 

 this character resembling Blachia, Dimorphocalyx, and a few allied 

 genera, which, however, have quite a different andrcecium. This 

 character, it may be remarked, has been used, e. g. by Sir Joseph 

 Hooker in the Flora of British hidia (vol. v. pp. 242-3), as a 

 means of separating the genera presenting it from others, such 

 as Trigonostemon, which have not an accrescent calyx. 



In a recently published part of Das Pflanzenreich (iv. 147, iii. 

 EtiphorbiacecB-Cluytiecs), Dr. Pax divides into four subtribes, based 

 primarily upon the nature of the andro^cium, the genera, with few 

 exceptions, included by Bentham & Hooker {Genera Plantarum, 

 iii., part i., pp. 248-9) in the subtribe Grozophorece.' These sub- 

 tribes constitute the tribe Gluytiea. Trigonostemon is placed in 

 the subtribe GluytiincB, among genera with definite stamens, while 

 to Blachia and its more immediate allies, with their indefinite 

 stamens, are assigned positions in the subtribe Godiceince. From 

 the clavis of this last subtribe we learn that Dr. Pax follows his 

 predecessors in dividing the genera into two groups, namely, those 

 with a non-accrescent and those with an accrescent calyx. Seeing 

 this, I naturally concluded that he would have recognized precisely 

 the same difference between Trigonostemon and Nepenthandra as 

 of generic validity ; but to my surprise found the new genus merged 

 in the old without the essential distinction between them being 

 assigned even sectional importance ! 



Dr. Pax refers to the plant in question in the following note : — 

 " Species calyce ? post anthesin accrescente valde insignis, sed 

 me judicante melius pro Trigonostemonis specie habenda quam pro 

 typo generis proprii " [l. c. p. 93). Whatever may have been the 

 practice in earlier days, it certainly is unusual in our time to 



