1916] Kroehcr: Floral Relations Among the Galapagos Islanfls 213 



So, also, compare Seymour twenty-six, and Jervis twenty-five — Ixitli in 

 James-Indefatigable waters — with Gardner twenty-three, and Barring- 

 ton eighteen, one in and tlie other near the southeastern province. 



Abingdon clearly lias some direct affinity with the two other north- 

 ern islands; l)nt this is less marked than might be surmised, until one 

 remembers that the three northern islands are rather small and not 

 closely grouped, so that in the long run the chances would be more 

 favorable of their receiving species from the large islands of the mass 

 of the archipelago than from one another. Even if the Galapagos are 

 not risen volcanoes but a gradually sunken land-mass, distinct local 

 species must have been often communicated from one island to another; 

 so that the point would hold. Proximity to Abingdon has, however, 

 had some influence in shaping the floras of Bindloe and Tower, as will 

 be shown ; liut on the other hand they are both too poor to have affected 

 Abingdon appreciably. 



DUNCAN 



Duncan, considering its size, has a remarkably varied flora, due 

 perhaps to the comparative variety of environment afforded by its 

 unusual altitude. It lies between Albemarle and Indefatigable and 

 near James. Its affinities are distributed about as one might expect 

 from its position and the relative wealth of species of the other islands, 

 except that the figure for Albemarle — seventy-five as compared with 

 seventy-five and sixty-nine for distant Charles and Chatham — sinks 

 rather low, and that for Narborough is surprisingly small. Evidently 

 Duncan has not been stocked in any great measure from the west, and 

 is itself too small to have had much influence on the larger western 

 islands. 



NARBOROUGH 



The unusual position of this island is of interest. It is the mo.st 

 w^esterly of the Galapagos, and is half surrounded, and shut off from 

 all the remainder of the archipelago, by crescent-shaped Albemarle. Of 

 the large islands, it is distinctly the poorest in flora, according to our 

 data. A large part of its area is covered by recent lava flows. It 

 might therefore be anticipated that Narborough would show a very 

 high degree of dependence on Albemarle, and little except the most 

 general- relationship to the other islands. This is only partially true, 

 86 per cent of its species are found on Albemarle; but this ratio is 

 substantially equalled by tlie 84 per cent of Hood species occurring 



