GEOGRAPHY OF PLANTS. 273 



some Scitamineae, rise to the same elevation. At the height 

 of 14,760 feet, we still find the Wax Palms, some Cinchonae, 

 Winterse, Escalloniae, Espelettiae, Culcitia, Joanneae, Vallea 

 stipularis, Bolax aretioides^ and some othei's. 



399. 



The growth of plants in society, or as individuals, is very 

 interesting. Many forms are so appropriated to certain re- 

 gions, that amidst constant changes they still take in a great 

 tract of land, and are produced in exuberant abundance. 

 Others, on the contrary, stand quite insulated, and seem as if 

 they would utterly disappear, did not Nature, in a manner 

 which is often inexplicable, provide for their continuance. ' 



While with us. Polygonum aviculare^ Erica vulgaris^ 

 Poa annua^ Aira canescens, Vaccinium Myrtillus, grow al- 

 ways in society, and cover great tracts of country, we ob- 

 serve, on the contrary, that Marruhium 'peregrinum^ Car- 

 duus cyanoides^ IStellera Passerina^ Carex Buxbaumii, Cir* 

 sium eriophorum, Lathyrus Nissolia, Hypericum Kohlianum^ 

 Schoenus ferrugineus^ and Helianthemum Furnana^ are con- 

 fined, in an insulated state, within a very narrow space, be- 

 yond which they never pass. The Cedar of Lebanon, Fors- 

 tera sedifolia of New Zealand, Melastoma setosum on the 

 Volcano of Guadaloupe, and Disa longicornis on some spots 

 of the Table Mountain of the Cape, are examples of this 

 completely insulated growth, which renders the idea of the 

 migration of plants at least very doubtful. 



400. 



If we proceed through the separate families, we shall find 

 their geographical distribution pretty exactly ascertained, and 

 their increase or diminution determined according to the dif- 

 ferent zones. But, as many families consist but of single 

 groups, which are limited to certain zones or countries, — this 

 circumstance occasions always a variation in the account. If 

 we attend, for instance, to the Rubiaceae, it is almost impos- 

 sible to pronounce any general opinion respecting the geogra- 

 phical distribution of this family, because the first group of 



S 



