200 University of California Publications in Botany [Vol. 6 



The most noteworthy feature of these differences is not, however, their 

 extent, but rather the fact that for the most part they stand in no relation 

 to the distance of the islands from each other or to the depth of the interven- 

 ing channels. Thus the florulae ot Albemarle and Chatham at opposite sides 

 of the archipelago are more alike than either is to that of the intervening 

 Indefatigable; Jervis lying near James has a greater common floral element 

 with the more distant Charles; the florulae of the Seymour Islands have a 

 greater number of plants in common with Charles, Chatham, and Albemarle than 

 with Indefatigable, of which the Seymour Islands are merely a detached spur. 

 Although a high percentage of ferns has been recorded on James, not a single 

 representative of this group has thus far been found on the adjacent Indefatig- 

 able. The common floral element between Duncan and the relatively remote 

 Chatham is greater than with any of the three large islands, James, Indefatig- 

 able, and Albemarle, which to a considerable extent surround it. In fact, 

 the only cases in which it appears that proximity between two islands has 

 brought about any marked similarity in their floras are on the one hand 

 Narborough and Albemarle, and on the other Gardner and Charles, and even 

 in the former of these pairs, the likeness is by no means close, for not over 

 71 per cent of the plants of Narborough have been observed on Albemarle. 



These anomalies in the different florulae must find their explanation in 

 peculiarities of climate and soil, together with an element of chance^ — arising 

 partly from imperfect exploration, and partly from the accidents of seed- 

 dispersal. Although they are not fully explained by the theory that these 

 are islands of emergence casually seeded, they are much less in accord with 

 the Baur theory of subsidence; for, were the florulae remnants of a common 

 flora persisting upon islands separated by gradual subsidence, it is evident 

 that those islands would possess the most floral similarity which were nearest 

 together and divided by the shallowest channels, since these would have been 

 sejjarated from each other more recently than the remoter islands, which are 

 cut off by a greater depth of ocean. As we find no such relation prevailing 

 in the Galapagos Islands, but have observed just the reverse, namely, that 

 the more distant islands, separated by relatively deep channels, often show 

 greater floral similarity than the nearer ones, it is necessary to conclude that 

 the botanical evidence, so far as it has been made out, is opposed rather than 

 favorable to the subsidence theory.2 



On the respective hypotheses of subsidence and emergence of the 

 Galapagos group, I have no particular conviction and no view to press. 

 But the basis of Professor Robinson's inferences on this point seems 

 questionable ; and a re-examination of the facts presented by him has 

 led me to a very different opinion of the irregularity of the inter-island 

 floral relations of the Galapagos. 



The statistical summary on which Professor Robinson bases most 

 of his conclusions is the following table of species and forms common 

 to the several islands.^ 



2 Ihid., pp. 244-259. 



3 Ihid., p. 253. 



