250 Dr. J. D, Hooker on the Vegetation 



the species and genera of those orders contained in one group of islands having 

 little or no relation with those of the others. 2ndly, That the chief points of 

 difference are explicable, and owing chiefly to the relations the islands bear to 

 the nearest continents, and to the nature of the soil and climate, &c. ; such 

 are the absence of Ferns, and the peculiar forms of Compositoe and Rubiacece, 

 and other orders having their nearest allies on the neighbouring continents, 

 3rdly, The smallest amount of novelty will be found amongst the more per- 

 fect plants, if such be so considered as possess a double floral envelope and 

 polypetalous corolla, including the Thalamifloroe and Legumlnosce, whilst the 

 greatest amount of new species exists in the lower orders, as Amaranthacece 

 and Piperacece, or in the incomplete genera of Euphorbiacece, and in the 

 Compositce. On the other hand, there are somewhat fewer peculiar and new 

 plants amongst the Monocotyledones than the Dicotyledones ; and the amount 

 of novelty amongst the Ferns is small in comparison to the higher orders. 



II. In this second part of the essay I propose to treat of the Flora of the 

 Galapagos as divisible into two types : these are the West Indian (including 

 Panama), to which the plants common to other countries and the dubious 

 species almost universally belong ; and the Mexican and temperate American 

 type, or that under which the great majority of the peculiar species will rank. 

 Those which I have called dubious species consist of a few plants which more 

 properly belong to neither of these divisions, including all such as are so 

 nearly allied to continental forms as to appear scarcely distinguishable specifi- 

 cally, or if distinguishable, dependent on characters which, though sufficiently 

 obvious, are extremely liable to variation ; such are possibly altered forms of. 

 introduced species, dependent on the combination of circumstances which they 

 are exposed to in the Galapagos for the appearances they assume : such plants 

 are noticed in the descriptions given of the species, and it is suflicient to state 

 here, that they nearly all belong to the West Indian type. 



The species which I have referred to the Mexican type (from the aflSnities of 

 the remarkable Compositce) include those whose nearest allies belong to Mexico 

 or the higher levels in Columbia, or to the lower latitudes of the Southern 

 United States, California or Chili ; unlike those of the West Indian type, they 

 are all specifically entirely distinct from their continental congeners, and are 

 about 45 in number, belonging to such genera as Discaria, Dalea, P/iaca, 



