Schow's Essay on Botanical Geography. 375 



and several arborescent Syngenesiw. Still much less resemblance is there 

 to the extra- tropical parts of New Holland. The likeness is stronger to 

 the tropical part of that country, as its Flora also approaches that of India. 

 Single genera only it seems to have in common with America ; for in- 

 stance, Melicocta, Ruizia, Dodonwa, Dichondra. These are probably pe- 

 culiar ; Latania, Hubertia, Poupartia, Tristemma, Fissilia, Cordyline, 

 Assonia, Fernelia, Lubinia, and others. Madagascar appears to possess a 

 very peculiar Flora ; it agrees most with the last mentioned islands, and 

 several genera are only found on them and Madagascar ; for example, Da- 

 mis, Ambora, Dombeja, Dufnurea, Didymeks, Senacea; several species 

 also are common to both, as Didymeles Madagascariensis, Danais fra- 

 grant, Cichona afro-inda ; but, however, among 161 known genera from 

 Madagascar, 54 only are found in the islands of France and Bourbon ; so 

 that there might be good grounds for forming a separate region of the first, 

 unless, perhaps, the east coast of Africa might come under the same. With 

 New Holland and the Cape, Madagascar has still less resemblance than 

 the two other East African islands. ( 



XVIII. Regnum Scitaminearum, (the Indian Flora.) To this belong 

 India, enst and west of the Ganges, together with the islands between 

 India and New Holland, perhaps also that part of New Holland which 

 lies within the tropics. The Scitaminea are here far more numerous than 

 in America, likewise, although in a less degree, Leguminosce, Cucurbitacew , 

 Tiliacece ; the before mentioned South American forms are more seldom, 

 or .else wanting. This region should certainly be divided into several pro- 

 vinces, but, as yet, we know too little of it, for us to undertake such a di- 

 vision with any degree of certainty. 



XIX. The Indian Highlands ought to have one, or probably two re- 

 gions, with a vegetation different from that of the Lowlands ; in the mid- 

 dle region, Melastoma, Orchidew, and Filiees, appear to prevail ; in the 

 higher, the vegetation approaches very near the European and North 

 Asiatic, and partly the Japanese; these districts probably form one region 

 with the whole of central Asia. 



XX. Of Cochinchina's and the]South of China's Flora, it partly comes 

 near India's, especially in regard to families ; but, however, Loureiro's 

 Flora contains a very large number of peculiar Genera. It is true, that 

 many of these genera ought probably to be reduced ; but, after all, the 

 vegetation of this tract will most likely be sufficiently peculiar for it to 

 form an individual region. 



XXI. The Flora of Arabia and Persia, seems likewise, with good 

 reason, to deserve to be separated from that of India, as it is already suffi- 

 ciently separated from the Mediterranean region (III.); for, amongst 281 

 genera in Fortkal, 109 only are found in the south of Europe; more like- 

 ly dots the Flora of Nubia, and part of central Africa, belong to this re- 

 gion. With regard to the numerous species ol* Cassia, and gummifcrous 



