ECOLOGY OF ISLE ROYALE. 179 



posed almost entirely of true Florldian plants. There are further vast stretches 

 of what is called 'the prairie,' i. e. land quite recently formed, partly by the-accumu- 

 lation of seaweeds swept ashore by the waves, and partly by the advance of the 

 Mangroves. This prairie is covered with the same herbaceous vegetation which 

 we see in similar places in Central Florida and does not contain a single semi- 

 tropical plant. Even the hammock is Invaded by several Floridlan trees: the 

 Live Oaks, several Palmettos, the Hackberry and others make their appearance 

 and, on higher ground we find plenty of Persea caroUnensis. Now on all these trees 

 in the pine woods and on the prairie, in short wherever there is the Floridian 

 flora we meet the true Floridlan insect fauna whereas the semitropical fauna is 

 confined to the semitropical forest.* This fact once recognized, it becomes evi- 

 dent that the northward extent of this fauna is identical with that of the semi- 

 tropical forest, a fact fully borne out by subsequent experience." pp. 170-171. 



"I desire to emphasize here once more as one of the principal characteristics 

 of this flora and fauna, that north of the Everglades they nowhere appear inland 

 but always close to the shore. Even along the inner bank of the Indian River 

 there are— or rather were — but a very few spots covered with semitropical forest, 

 viz: on the mouth of the St. Lucie and Sebastian Rivers, at the southern end of 

 Merritt's Island and perhaps some others; but they are now mostly destroyed by 

 cultivation." p. 172. 



1890. pp. 186-187. 

 ■ "The mountain ranges in America run in the direction from north to south, 

 and the colonies of circumpolar insects upon their summits have thus been able 

 to preserve their connection and specific identity with the arctic forms; whereas 

 In Europe, where the mountain ranges run from east to west, the alpine colonies 

 have generally undergone changes and, by isolation, lost their specific identity 

 with the arctic species. There is, therefore, in the Old World an abundance of 

 distinct alpine forms, none of which are identical with North American species; 

 while we,_on our high mountains, have but few, if any, alpine, but more arctic 

 forms, pp. 186-187. 



"Among the strictly circumpolar Coleoptera the predaceous families predom- 

 inate over the phytophagous families; the Carabidae, Dytiscidae, Staphylinidae, 

 and CoccinelUdae are well represented, the Chrysomelidae and Rhynchophera are 

 tolerably well, and the Cerambycidae and Elateridae are poorly represented. The 

 Buprestidae are absent although this family contains numerous boreal species 

 in every region. The phytophagous Scarabaeidae do not, or barely extend into 

 the arctic regions; the coprophagous Scarabaeidae (Aphodius) are well repre- 

 sented there, still none of them (with the exception of Aphodius rufipes, which 

 doubtfully belongs here) is on the list of circumpolar Coleoptera." p. 187. 



"Species not Belonging to the Circumpolar Fauna. — This division comprises 

 endemic species of probably intratropical origin, which have spread, by natural 

 dispersion, into the temperate zone of North America." p. 187. 



1890a. pp. 170-171. 



"Turning now to the bulk of the species in the list [St. Augustine, Florida] we 

 find that they consist of the usual admixture of more or less widely-distributed 

 species and true Florldian forms, the proportion being but little different from that 

 of the other localities, e. g.. Crescent City, Enterprise, Tampa. . . But the St. 

 Augustine list contains another element, viz: species belonging to the faunal 

 region lying directly north of eastern Florida and comprising lower Georgia, the 

 lower Carolinas, and eastern Virginia. This is an ill-defined region with very 

 lew, or no, peculiar species, and only characterized by a certain combination of 

 a number of southern species. The existence of this faunal region will become 

 evident to any one who, on a summer day, goes from here [Washington] down 

 to Fortress Monroe, Va. The difference between the Washington fauna and that 

 of Fortress Monroe will then be found quite striking. Of this fauna I noticed 

 about twenty species in the St. Augustine list not previously known from Florida." 

 pp. 170-171. 



1901. pp. 1, 2, 3. 



"Still, southwestern Texas belongs, at least as far as the insects are concerned, 

 to the lower Sonoran fauna, of which it forms a marked subdivision, but with 

 marked affinities to the austroriparian region. 



• " * There is, in addition, in Southern Florida a maritime fauna of semitropical character, but the 

 number of species composing the same (about 12 in Coleoptera) is so small that it is hardly worth 

 while considering. Its northern extent is still uncertain but it is safe to say that on the eastern coast 

 it does not reach beyond Mosquito Inlet at New Smyrna." 



