246 



MEMOIRS OF THE NEW YORK BOTANICAL GARDEN [Vol. 8, No. 3 



Crotalaria aculeata De Wild. Ann. Mus. Congo Bot. IV. 185. pi 46, f. 28-28. 1903; 

 Verdoorn, Bothalia 2: 413- 1928. 



Crotalaria spinosa [non Hochst. ex Benth.l Benth. Lond. Jour. Bot. 2: 576. 1843, pro 



parte, quoad spec. Kotschy 552. 

 Crotalaria spinosa Hochst. ex Benth. var. pubescens Benth. Lond. Jour. Bot. 2: 576. 



1843. 



Ononis emarginata Boj. ex Benth. Lond. Jour. Bot. 2: 576. 1843, pro syn. 

 Crotalaria spinosa Hochst. ex Benth. subsp. aculeata (De Wild.) Bak. f. Jour. Linn. 

 Soc. Bot. 42: 312. 1914; Leg. Trop. Afr. 37. 1926. 



Chikwawa District: Lower Mwanza River, scattered on sandy beaches, herb 60 

 cm. high, flowers yellow, pods inflated, 180 m., Oct. 6, 1946, 18020. Anglo-Egyp- 

 tian Sudan, Belgian Congo, Uganda, Tanganyika Territory, Nyasaland, N. Rhode- 

 sia, Angola, and Madagascar. 



This is a member of a taxonomically most difficult complex. According to E. 

 G. Baker's view C. aculeata is a subspecies of C. spinosa Hochst. (ex Benth. 

 Lond. Jour. Bot. 2: 576. 1843); with this I disagree, since I am quite unable to 

 find those intermediates that, according to Baker (Jour. Linn. Soc. Bot. 42: 312. 

 1914), connect the two. Baker was perhaps influenced by the fact that Bentham 

 (I.e.) cited two specimens under C. spinosa — Schimper J 50, which is the type- 

 number, and Kotschy 552, which is certainly C. aculeata De Wild. 



True C. spinosa is distinguished from all forms of C. aculeata, firstly by its 

 very small flowers 4-6 mm. long — the distance from the bend in the keel to the 

 tip is only 4-4.5 mm.; and secondly by the small pods 6.5-8(-9) mm. long and 

 3.5-5 mm. in diameter. The calyx is only 2.5-4 mm. long with lobes 1.5-2 mm. long. 



In a single specimen only, Rounce 52 (Tanganyika Territory, Kasulu), lacking 

 ripe pods, the length from the bend in the keel to the tip is up to 6 mm. long, but 

 even this is less than the shortest-flowered C. aculeata. 



True C. spinosa is found in Socotra, Eritrea, Abyssinia, Uganda, Kenya, and 

 Tanganyika Territory with an isolated occurrence in Angola (Welwitsch 1908, in 

 Herb. Kew), and on the whole seems a constant and distinct plant. It appears to 

 be absent from Madagascar, but occurs in Socotra. 



With C. aculeata the story is very different. It is easily separable from C. 

 spinosa by its much larger flowers and pods. The corolla is 9-10 or occasionally 

 up to 20 mm. long, and the distance from the bend in the keel to its tip is 8-10 

 (-15) mm. The pods vary between 9 and 30 mm. in length, and 5 and 15 mm. in 

 width. 



The commonest and smaller-flowered forms of C. aculeata have corollas about 

 9-10 mm. long, and the length of the keel from bend to tip 8-10 mm.; the calyx is 

 4.5-5 mm. long with lobes 1.5-2.5 mm. long. Flowers of this size can produce 

 pods ranging from 9-21 mm. in length and 5-9 mm. in width. C. claessensii De 

 Wild. (Bull. Jard. Bot. Brux. 3: 271. 1911; Ann. Mus. Congo Bot. V. 3: 410. 1912) 

 seems to me only a form with pods near the upper limit of size. 



The common small-flowered forms of C. aculeata range from the Anglo-Egyp- 

 tian Sudan southwards to Angola. In the southern part of this area, particularly N. 

 Rhodesia and the southern Congo, forms with larger corollas occur. These are in- 

 constant and most perplexing. Mrs. Macaulay 627, from Mumbwa in N. Rhodesia 

 is an extreme with flowers up to 20 mm. long, and the keel 14-15 mm. from bend 

 to tip; the pod reaches about 30 mm. in length and about 10 mm. in breadth. This 

 plant is the type of C. spinosa var. macrocarpa Bak. f. The description of C. 

 kapiriensis De Wild. (Bull. Jard. Bot. Brux. 5:23- 1915) reads so similarly that I 

 feel it must be the same thing. But these are extremes. In /. D. Martin 610 from N. 

 Rhodesia the pods are only 20-25 mm. long, and specimens such as Quarrk 3161, 

 Milne -Redhead 730, and Kassner 2713 show gradual reduction in flower-size to 

 that of normal C. aculeata. 



