vi DISTRIBUTION AND EVOLUTION 97 



Florissant, Colorado, of middle or late Tertiary age, which 

 shows signs of a much milder climate than now prevails 

 there. Many of these plants are of genera now extinct or 

 only found in more southern lands, and this extinction is 

 traceable to the great changes, inorganic and organic, that 

 have since occurred in North America. He says (in a private 

 letter) : 



" There was first the invasion of Old World species via Behring's 

 Straits ; then an incursion of S. American forms vid Panama ; and 

 then the glacial period at the end, crowding and destroying much 

 of the flora and fauna. Since the glacial period in N. America, 

 there has been room for expansion, and hence the very numerous 

 and closely allied species of Aster, Solidago, Senecio, and other 

 plants, as well as allied species of butterflies of the genera Argynnis, 

 Colias, etc. These are, most of them, not at all on the same footing 

 as the tropical species. ... I think tropical species are better 

 defined than those of the temperate region." 



It is a rather curious coincidence that if we take the 

 mean area of the twelve English counties for which I have been 

 able to give the figures, in geographical instead of English 

 miles, the number of square miles will almost exactly equal 

 the average number of their species of flowering plants. 

 Below this area, in the mid-temperate zone, the proportion 

 of species decreases, and above it increases, in both slowly 

 at first and with many fluctuations, but afterwards very 

 rapidly, more especially for the larger areas, so that it requires 

 on a rough average about a two hundred-fold increase of area 

 to double the number of species^ and about a thousand-fold 

 to quintuple it. But in all such comparisons we require a 

 large number of fairly comparable cases to give a trust- 

 worthy average, and the materials for this do not seem to 

 exist. Yet there is a striking general agreement between 

 the numbers of the species in the various kingdoms, states, or 

 colonies of Europe, North America, and Australia, requiring 

 only slight allowances for greater area, better climate, or 

 geological history to bring them into line with one another 

 to a really remarkable degree. 



It appears, then, that, whether we take small areas 

 roughly approximating 100 square miles, or much larger 

 areas of from 100,000 to 200,000 square miles, there is, 



H 



