336 THE JOURNAL OF BOTANY 



served in the Smithian Herbarium, marked (by Buchanan, as it 

 seems) as " G. excelsaV 



It will be noted that Eoxburgh did not profess to have any 

 direct knowledge of this species, and the description is a word for 

 word translation from the Symbola, but he had in fact sent a 

 specimen of G. vestita Wall, to Banks as No. 98 of the set of 

 specimens transmitted (1791 to 1794) to Banks with the figures 

 and descriptions of five hundred Indian plants mentioned by King 

 in his Brief Memoir, of which someone, probably Dryander, selected 

 three hundred, which were published in Coromandel Plants. It 

 should be added that in no case can either the plant issued by 

 WalHch as vestita, or that described at p. 104 of Royle's Illustra- 

 tions as elastica be referred to G. asiatica Linn., or to G. suh- 

 incequalis DC, if the latter be the Calcutta Garden ''asiatica." 

 So far as the writer has been able to form an opinion, both 

 vestita and elastica are simply climatic races of a type extend- 

 ing from the Indus Valley as far east perhaps as the Philip- 

 pines, of which Jussieu's celtidifolia represents the chief Malayan 

 form. 



So far nothing has been said of the ''G. arborea" of Roth, 

 which he received from Heyne under the name of ''arborea Rox- 

 burgh." Heyne's examples under the like title in Rottler's her- 

 barium are all typical tilicefolia, but Roth's description indicates 

 G. rotundifolia Juss., and does not answer to tiUafolia Vahl. It is 

 noteworthy that rotundifolia, unless traceable under some other 

 description, is not accounted for in the Flora Indica, and the 

 probability is that Roxburgh and the missionaries, in the beginning 

 at least, regarded G. rotundifolia, the true tilicefolia, and certain 

 other forms of which, for the present, we know very little, as 

 " varieties " of a single polymorphic species, for which, as the 

 first studied example was a good -sized tree, Roxburgh pro- 

 posed the name "arborea," in contradistinction to the often 

 scandent shrubs, such as hirsuta, with which he had been earlier 

 familiar. 



Let us now return to species No. 8 {" salvifolia"),'Q.nd consider 

 in connection with that species Nos. 9 (" hirsuta"), 11 {" carjnni- 

 folia"), 12 {"lyilosa"), and 13 {" ]Jolygama"), which belong to one 

 natural group, whereas " salvifolia " finds its affinities in the 

 Tilicefoliai, and particularly with rotundifolia Juss., from which 

 in some states it is hard to distinguish it. 



We have seen already that the original G-, salvifolia of the 

 younger Linne, so far as the description goes, was not any Greioia 

 at all. There is some reason to suppose that Alangium liexa- 

 jJetalum Willd. was sent to Europe mixed with G. ftavescens Juss. 

 {= carjnnifolia Roxb. non Juss., also = commutata DC), both 

 being labelled by the senders " G. montana" (of the missionaries). 

 British Museum specimens, gathered by Koenig or a colleague 

 "ad latera et summitates montium," show that the type of 

 " G. montana" was the plant which Vahl has described as G. hir- 

 suta. The name "montana" was dropped, probably with regard 

 to the younger Linn6's " salvifolia," but in the meantime Rox- 



