292 XLVIII. SAPOTACE.E. [Mimusoj 



culty. Linnaeus in his Flora Zeylanica (1747), p. 57, describes two species : 

 M. Elengi (foliis alternis remotis), regarding which there is no doubt ; and 

 another, subsequently, in Sp. Plant., called by him M. Kauki (foliis confertis). 

 These species were based upon specimens collected in Ceylon by Paul Hermann, 

 Professor of Botany at Leyden, between 1670 and 1677, and now preserved in 

 the British Museum. Robert Brown (Prodr. Fl. Novae Holl. 1810, p. 531) iden- 

 tifies with M. Kauki an Australian tree, which had been found by Solander in 

 1770 on islands off Cape Fear in Queensland. Grisebach (West Ind. Fl. 1864, 

 p. 400) adheres to this view, and defines the area of the species as " Pacific 

 islands, tropical Australia, and the East Indies on the sea-shore." He also refers 

 to it a tree cultivated in the West Indies, and figured by Sir Wm. Hooker in Bot. 

 Mag. t. 3157, under the name of M. dissecta. Bentham, however (Fl. Austr. iv. 

 285), considers that Hermann's Cingalese specimen differs from the Australian 

 plant, which he calls M. Browniana, and that it should be referred to M. 

 indica, A. DC, a tree, with extremely hard strong and very durable timber, 

 which is common in the hot, drier parts of Ceylon (Thwaites Enum. 175). I 

 have also examined Hermann's specimen, and though I do not venture to offer 

 any opinion regarding its identity with the tree from Australia and the Indian 

 Archipelago, I have no hesitation in saying that it cannot be referred to the 

 North Indian tree. The leaves are obovate, acute, and the petiole more than 

 one- half the length of the blade. They are pale beneath, though not in so 

 marked a manner as the leaves of the tree from Australia and the Indian Archi- 

 pelago ; but it must be remembered that the specimen is 200 years old. It is in 

 bud, too young for a satisfactory examination of the parts of the flower. The 

 North Indian tree has concolor leaves, which are obovate-oblong, obtuse or 

 emarginate, with a petiole about J the length of leaf. The figure of Rumph. 

 Herb. Amb. iii. t. 8, quoted in Linn. Spec. Plant., does not prove anything either 

 way ; and under these circumstances I do not feel warranted in adopting the Lin- 

 naean name for our tree. The Linn. Herb, at the Linnaean Society contains two 

 specimens of Mimusops : one is M. Elengi, though marked Kauki, Konig, in Lin- 

 naeus' handwriting (apparently by way of indicating the origin of the specimen), 

 but Elengi in pencil by Sir J. E. Smith ; the other is very incomplete, and may 

 be a large-leaved form of M. Elengi, it certainly is not Khirni. It is marked 

 M. Elengi in Linnaeus' handwriting, and M. Kauki ? in that of Sir J. E. 

 Smith in pencil. Roxburgh's M. hexandra is supported by t. 15 of the Coro- 

 mandel plants, and by a type specimen, in leaf only, in the Wallichian herbar- 

 ium. This specimen apparently belongs to the tree under discussion, and Rox- 

 burgh's description also is evidently intended for it. But as pointed out in DC. 

 Prodr. viii. 204, and in Wight's 111. ii. p. 144, the staminodes are not correctly 

 represented in the plate ; and it is possible, as Wight suggests, that the flowering 

 branch and fruit was taken from one species (M. indica), and the magnified 

 flower from the other (if. Roxburghiana). Roxburgh's name, therefore, cannot 

 be admitted. It is, moreover, inappropriate, as the tree has both hexandrous and 

 octandrous flowers. Nothing therefore remains but to fall back upon De Can- 

 dolle's name, M. indica, which is supported by good descriptions and the plate 

 in Wight's Icones. The specimens of the tree from South and North India are 

 identical, though there is a certain amount of variation in the shape of the stam- 

 inodes, which, however, in all Indian specimens examined by me, and in some of 

 the Ceylon specimens, are bifid. There are, however, Ceylon specimens with 

 entire, not bifid staminodes, longer than stamens, which may possibly belong to 

 a different species. 



S. Kurz, in his Report on the Vegetation of the Andamans, and in Journ. As. 

 Soc. xl. 1871, p. 70, calls the Andaman Bullet-wood, M. indica. The Martaban 

 tree, which is (probably) erroneously quoted by De Candolle under this species, 

 may very likely be identical with the Andaman Bullet- wood, but the Hindustan 



