THE GREWIAS OF ROXBURGH 337 



burgh had sent to Smith as " G. orientalis " (which it was not) a 

 gathering now in Smith's herbarium, which Smith, remarking 

 that it was the G. orientalis of Vahl and Gaertner, referred to 

 G. pilosa of Lamarck (Encycl. iii. 43, ann. 1789). There is a 

 scrap of the same plant in the Linnean herbarium, not named by 

 either Linne, which is also written up by Smith as " G. j^Hosa," 

 but erroneously, both specimens being really G. carplnifolia Eoxb. 

 non Juss. — i. e. G. flavescens Juss. — (syn. G. commutata DC. Prodr. 

 i. 511). This error of Smith's has led to the subsequent very 

 general misapplication of " pilosa Lamarck" to G. flavescens, hoth. 

 in Asia and Africa. This was not all, however, for on a specimen 

 collected in Java, possibly by Thunberg, which an unknown 

 annotator has correctly identified as G. toinentosa Juss., Smith 

 made a note referring it to hirsuta Vahl, " ex characteris." All 

 this helps in some measure to account for two puzzling circum- 

 stances, viz. (a) why Roxburgh, having in the first place named 

 his Coromandel " JovelUkee " (No. 9 of the Flora Indica) " tonien- 

 tosa" in MS., presently cancelled this in favour of ''hirsuta" 

 Vahl ; ih) why Smith (in Bees Cijclopadia, xvii. s. v. Greivia), and 

 later A. De Candolle (Prodr. i. 509), have given Java as the home 

 of G. hirsuta Vahl (c/. Wight & Arnott, Prodr. 78, footnote). In 

 point of fact the Javan G. toinentosa, which is found in Tonkin as 

 w'ell as in Java, but has not so far been collected further to the 

 westward, is abundantly distinct from the true hirstita Vahl, which 

 conversely is a purely Indian type ; for the forms referred to it 

 from the Indo-Chinese region by Pierre [Flore Forestiere, v. 

 pi. 166 and 167, and text) are distinct from true hirsuta. 



What happened apparently was more or less as follows. In 

 the first instance Eoxburgh supposed that he had traced what 

 should now be called G. Eothii DC. to the description in the 

 Su-pplementum by the younger Linne of a more or less imaginary 

 Grewia, which as regards the material actually sent from India 

 was at least in part an Alangium. G. Bothii was apparently not 

 firmly discriminated by Koenig and his friends from the allied 

 form first described by Gaertner from Ceylon material as G. Daviine 

 (De Fruct. ii. 112, t. 106, ann. 1788), so that in the end it was 

 G. Damine that was sent as "G. salvifolia" to Roth by Heyne, 

 and published as " G. salvifolia" by Roth (Nov. Plant. Spec. 239 



(1821)). 



We have seen, however, that the missionaries' " G. montana" 

 probably included one or more examples of a true Greioia, viz. 

 G. commutata of De Candolle, which is = G. flavescens Juss., a 

 species ranging from Senegambia to the Coromandel coast of 

 India in the drier regions of the subtropics. This species was 

 one of the earliest collected in South India, and had been regarded 

 as a variety of G. orientalis of the Sj^ecies Plantarnm ; when it 

 came under Smith's eye he ascribed it to the G. pilosa of the 

 Encyclop&die. 



(To be conlimieil.) 



