of the Galapagos Archipelago. 247 



form, thus : the more an island is indebted to a neighbouring continent for its 

 vegetation, the more fragmentary does its flora appear, migration being efifected 

 by the transport of isolated individuals, generally in nowise related, while an 

 independent flora is generally made up of groups, the lowest order of which 

 we call genera. Hence the Coral Islands, whose flora is wholly borrowed, sel- 

 dom have two species belonging to the same genus ; as Keeling Island, for 

 instance, on the west, and Maiden Island to the eastward of the Pacific, in 

 both of which the genera are to the species as 1 : 1 ; or the Society group, 

 which presents so little novelty, and whose proportion of genera is as 1 : 1*3. 

 In the Sandwich Islands, on the other hand, our very imperfect materials give a 

 proportion of 1 : 2 ; the Galapagos of 1 : 1*7 ; and St. Helena about the same. 



In the abundance and peculiarity of Rubiacece, the Galapagos may be com- 

 pared with the Sandwich group, which contain 18 species of this order unexcep- 

 tionably peculiar, and all belonging to genera even confined to those islands. 

 None of the Galapageian species belong to any but very common genera*. 



Of SoLANEiE there are 13 species, a considerable number in proportion to 

 the extent of the Flora, but almost the whole of them derived from the 

 neighbouring coast. Only one, a species of Acnistus, is previously unde- 

 scribed, except the Dictyocalyx, which is however a native of the shores of 

 Peru. Solanum nigrum is the only flowering plant common to Great Britain 

 and the Galapagos, except Ferhena officinalis and Calystegia Soldanella ; all 

 are found in most intervening shores. 



The Amaranthace^ are the only other family of flowering plants appearing 

 to demand a particular notice, for their abundance and novelty are without a 

 parallel in any tract of land of equal extent. This is partly to be accounted 

 for by their being partial to the immediate neighbourhood of the sea, but is 

 more due to the position of the Galapagos being near to America, where, 

 according to Martins, they attain their maximum. Littoral and widely dif- 

 fused as many plants of this order are, we do not find one of the ubiquitous 

 species in this group ; and indeed, except two belonging to the genus Ama'- 



* In respect of Rubiacece, the vegetation of the Sandwich group has no analogy with that of any 

 other islands. The Society Islands contain many more species, but all belonging to Asiatic genera. 

 Even the specifically peculiar Ruhiacex of St. Helena and Ascension have Cape congeners, and there 

 is no genus of this order confined to Juan Fernandez. 



