DERIVATION OF THE FLORA AND FAUNA 357 



Zealand's continentality. Accordin<^ to Ciikkskman the number of families is 97, 

 of genera 382 and of species 1415; the relation in question is 3.7: i. I cannot 

 attach much importance to such figures. If an island is a remnant of a sub- 

 merged land-mass upon which during the process of sinking la\a was ejected until 

 tiie old foundation disappeared, only a small part of the flora and fauna will sur- 

 \i\c the catastrophe, and unless progressive endemism comes to play a role, the 

 living world will present a picture of disharmony, with a reduced number of 

 families and a low species : genus ratio. Climatic changes will create a similar 

 situation, if a portion of a continental flora has lost its connection with the con- 

 tinent. The Falkland Islands rest on the continental shelf, they are formed by old 

 sediments on a granitic foundation and arc, universall}- I think, classified as con- 

 tinental islands, but their angiospermic flora is disharmonic and fragmentary: 38 

 families, 97 genera and 143 species; the ratio is 1.47 : i. There is a single endemic 

 genus and a small number of endemic species. The poverty is due to a severe 

 climate and to losses suffered during the pleistocene j)eriod of solifluction, con- 

 temporaneous with the glaciation on the mainland. The constitution of a flora, con- 

 tinental or insular, depends on a combination of many factors, geological, his- 

 torical, climatic, genetic and so forth. 



Absence of large, widespread and ^'successfuV families: — The lack of conifers 

 in islands demonstrates, it is said, that such islands are not continental, for in 

 all continents gymnosperms are plentiful; "cones do not float", and the seeds, 

 winged or unwinged, have no chances to be carried very far, but it is surprising 

 that also Taxads and many Podocarps, adapted, as it were, to endozoochorous 

 bird dispersal, are conspicuous by their absence. Consequently, islands where cone- 

 bearing species exist are regarded as continental: New Zealand, Tasmania, Nor- 

 folk Island, New Caledonia, Fiji, etc. 



Exceptions occur, islands looked upon as permanently isolated, Bermuda, 

 the Azores, Madeira and the Canaries, have at least some species oi Juniperus; 

 perhaps junipers possess some kind of dispersal capacity and do not count, but 

 we cannot get away from Finns canariensis. 



Other large and wide-spread families very poorly represented on isolated is- 

 lands are Leguminosae, Araceae and Orchidaceae. Peas are a staple food of many 

 birds but few if any able to pass their digestive tract unharmed; they are too 

 heavy to be carried any distance by wind and, notorious beach plants excepted, 

 will sink in water. The dry pods would float but they do not fall off but open 

 on the plant to discharge the seeds. Nevertheless Sophora sect. Edwardsia has a 

 number of closely related species scattered over the south hemisphere on islands 

 as remote as Diego Alvarez, Rapa, Easter and Marquesas, with one species, more 

 well-marked than the rest, on Hawaii. This is a very puzzling case of di.sjunct 

 distribution, but it is not expedient to draw any far-reaching conclusions from 

 this unique case. The seeds are of the size of a small pea and not equipped with 

 a capacity to travel greater than in hundreds of leguminous plants, which are 

 within reach but never crossed any water barriers, nor used any land bridges. 

 Did they not yet exist, were they not within reach when connection was estab- 

 lished or have they all died out in the islands? Flach alternative seems equally 



