LVTZ, LIST OF GREATER ANTILLEAN SPIDERS 131 



puts them ou three sides of the Antilles or, as is more probable, they are 

 species which can live in a wide range of environments, as is indicated 

 by their wide geographic range, and so have found it easy to estal^lish 

 themselves in the Antilles. 



The Central American affinity is really slight. Although in the totals 

 it seems to be important, an examination of the more detailed part of the 

 table shows that this is caused by the groups in which it is joined with 

 either the northern or the southern division or with both. For example, 

 the 17.2 per cent, of the mainland species in the Lesser Antilles which 

 are credited to South America and Central America should probably be 

 credited to South America as is indicated by the fact that 31.0 per cent, 

 are known on the mainland only from South America. The 5.2 per cent, 

 which are known from the mainland only in Central America are three 

 species M^hich may have come directly from Central America but they 

 more probably occur in South America or did recentl}^ occur there. 

 Jamaica is the only island with a greater percentage of species which, 

 as far as we know, are strictly Central American, than it has of species 

 which seem to be confined to one of the other mainland divisions, but 

 this percentage is misleading also as it represents but one species. 



Presumably these species have reached their island homes from the 

 mainland or have originated in the islands and spread to the mainland 

 since the various islands have been separated, if, indeed, the islands were 

 ever joined. The fauna of the islands at the ends of the Antillean chain 

 have their strongest affinities with that part of the mainland w^hich is 

 nearest to them partly because of their geographical contiguity but partly 

 also because of resemblance of habitat, witness the Floridian affinity of 

 the fauna (and it is true also of the flora) of the sandy plains near Pinar 

 del Eio, Cuba, as compared with the more tropical character of the moun- 

 tains to the north of the city. If these species have spread themselves 

 in this way without the aid of actual land connection, is such connection 

 indicated in the older units, the genera ? 



It is doubtless clear that if we take all genera and treat them as we 

 have just treated the species the result would be influenced by the in- 

 clusion of recently introduced species. Probabh^ — ^but certainly not nec- 

 essarily — those genera which have species peculiar to the Antilles have 

 been on the Antilles longer than those wdiose only Antillean species occur 

 also on the mainland. Table XI gives the percentages of such presum- 

 ably old genera occurring in a given island and also on the mainland 

 which are found in the several mainland divisions. It is an attempt to 

 discover the wav tlie fauna moved in relativelv remote times. 



