ORIGIN OF THE GALAPAGOS ISLANDS 199 



Scharff 17 with the statement that I have produced no new data 

 for or against a former land connection with the mainland. It 

 is true that I did not discuss this phase of the subject at great 

 length in my former paper on the Galapagos flora, because, 

 after I had tabulated my results 18 the evidence was so strongly 

 against the theory of a continental origin that I considered an 

 extended discussion superfluous. This table shows that of the 

 eighty-one families. of vascular plants represented upon the is- 

 lands, only thirteen contain ten or more species. Of these 

 Filices, and Compositae stand at the top of the list with seventy- 

 seven and sixty-nine species respectively. Such a condition 

 as this could hardly be possible if the islands were the remains 

 of an extension of the continent, because many of these families 

 are represented by a much larger number of genera on the main- 

 land at the present time. We might well ask the adherents of 

 a continental origin, why this condition should be, or why so 

 many families are represented by but a single genus or even 

 species. Surely the mere fact of isolation would not account 

 for it because approximately 60% of the species of plants on the 

 islands at the present time have a wider distribution on the 

 mainland. Such a large per cent of non-endemic species would 

 hardly remain if the conditions in the past had been very adverse. 

 It seems that this condition can hardly be accounted for unless 

 an oceanic origin of the islands is admitted. My work on the 

 botany of Cocos Island 19 strongly bears out this view, as I have 

 shown that the flora of this island differs greatly from that of 

 the Galapagos. If we exclude those species which are known 

 to have been introduced on both islands, there are but twenty- 

 four common to the two, eleven of which are ferns. There are 

 several more species of rather wide distribution common to the 

 two, which might easily have been introduced with garden seeds, 

 etc. In fact, Captain August Gissler, a resident of the Cocos 

 Island for many years, told me that Fleurya aestuans Gaud, 

 had appeared upon the island since he had been there, and 



17 L. c, pp. 315-316. 



18 Stewart, /. c, p. 235. 



19 Notes on the Botany of Cocos Island. Proc. Cal. Acad. Sci., fourth series, 

 Vol. 1, pp. 375-404, 1912. 



