244 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vo/. K 8, No. 3 



4473; ? without locality, Mar. 20, 1906, Gren/ell 14, 16. Barberton District: Avoca near 

 Barberton, 500 m. , July, 1931, Thorncroft 3060. Wolmaransstad District: Makwassie 

 Spruit near Wolmaransstad, c. 1700 m., Mar. 3, 1935, Liebenberg 3410. 



CAPE PROVINCE: Vryberg District: Vryberg, 1220 m., April 11, 1921, Mogg 8917. 



I am satisfied that the Transvaal specimens cited by Engler, Schonland and 

 other authors under the name R. gueizzii Sond. are specifically distinct from type- 

 material of that species (Port Natal, Gueinzius s.n. in Herb. Kew.). Schonland 

 (Bothalia 3: 79. 1930) has evidently confused a Gerrard specimen with the Guein- 

 zius type, for he refers to the latter as Gueinzius 1395, whereas the type is sine 

 numero, he also comments: "Drupe in type subglobose, verrucose. . . but the 

 type of R. gueinzii consists solely of three male flowering shoots. Gerrard 1395 

 is probably intended: this is a fruiting specimen intermediate in character be- 

 tween R. gueinzii and R. crispa (Engl.) Harv. ex. Schonl. I very much doubt if R. 

 crispa can be distinguished from R. gueinzii, even as a variety. The undulation 

 of the leaf-margin is a variable character, and is present to a greater or lesser 

 degree in both species. True R. gueinzii Sond. differs from R. amerina in having 

 the young twigs white or pale-grey, not brown or reddish; the leaflets are obtuse 

 at the apex, and turn black or blackish on drying. The fruits of R. gueinzii are 

 subglobose and ± verrucose, those of R. amerina are strongly compressed and 

 glossy. 



I have not seen specimens of R. gueinzii Sond. var. brevifoliolata Burtt-Davy 

 (Kew Bull. 1921: 51. 1921); from the description, the variety would appear to be 

 allied to R. amerina rather than to true R. gueinzii. 



Heeria reticulata (Bak. f.) Engl. Pflanzenw. Afr. 3 a : 197. 1921. 



Heeria insignis (Del.) Kuntze var. reticulata Bak. f. Jour. Bot. 37: 428. 1899. 



Kasungu District: Kasungu Hill, occasional on rocky slopes, tree about 8 m. 

 high, sap milky, fruit black, 1100 m., Aug. 28, 1946, 17453. Recorded from Portu- 

 guese East Africa, Nyasaland, N. and S. Rhodesia, Tanganyika, and (doubtfully) 

 Kenya. 



The proper status of Heeria reticulata is not easily decided. Extreme forms 

 with very prominent tertiary venation and dense woolly indumentum are distinct 

 enough, but as one travels north of the Rhodesias the differences between this 

 species and H. insignis (Del.) Kuntze are less apparent, and some of the Kenya 

 specimens might be referred with equal justification to either species. For the 

 present, I am content to follow Engler, and treat H. reticulata as a species; it is 

 clear that this particular problem is one which can be solved only with the co- 

 operation of the field-worker. 



Sorindeia madagascariensis Thou, ex DC. Prodr. 2: 80; Perrier de la Bathie in 

 Humbert, Fl. Madag. 114: 26. 1946. 



Cholo District: Nswadzi River, in riverine rain-forest, tree 10 m. high, pani- 

 cles numerous, pendent, on branchlets below the terminal cluster of leaves, 

 flowers orange-coloured, 840 m., Sept. 29, 1946, 17864. Kenya to Nyasaland and 

 Portuguese East Africa, also in Madagascar and the Mascarenes. 



I cannot see that S. obtusifoliolata Engl. Pflanzenw. Ost-Afr. C: 244 (1895) 

 differs specifically from S. madagascariensis. In Pflanzenw. Afr. 3 2 : 190 (1921), 

 Engler keys out S. obtusifoliolata by its oblong-obovate scarcely pointed leaf- 

 lets, but I find the shape very variable indeed both in Madagascar and in the 

 plants from continental Africa. 



Lannea edulis (Sond.) Engl, in Engl. & Prantl, Nat. Pflanzenf. Nachtr. 1: 213. 



1897. 



Odina edulis Sond. in Harv. & Sond. Fl. Cap. 1: 503. 1860. 



