CH. xvii] AND GENERA OF LARGER SIZE 193 



explainable by their aid. How could local adaptations be gradu- 

 ated in this regular order, or how could there be a vast number 

 at the last stage of relicdom, and fewer and fewer at the stages 

 leading up to that, and that in every family or country? 



Another great difficulty for the older explanations is provided 

 by the increase of monotypes, as of endemic genera and species, 

 as one goes southwards and outwards. Why should New Cale- 

 donia, the Mascarenes, and Juan Fernandez require so many 

 more per thousand square miles than the Sandwich Islands, 

 Formosa and Cuba, in similar northern latitudes, especially as 

 their non-endemic genera are in general very large and "suc- 

 cessful" world-ranging genera? AVhy should Chile have about 

 100 local monotypes, while there are only about 77 in Europe, 

 with more than ten times the area? Why should W^estern Asia 

 require so many more than Europe? and so on. 



The only reasonable explanation of the very striking facts that 

 have been set forth in the last three chapters, so far as I can at 

 present see, is that provided by Age and Area with its corollary 

 Size and Space, that the smaller genera are as a rule the younger, 

 that they are probably the descendants of the larger genera, 

 that they gradually increase their area with their age, and that 

 as the area increases, so does the number of species, these also 

 increasing their area with their age. As a general explanation 

 of the phenomena seen in the distribution of plants about the 

 globe, this commends itself by its extreme simplicity, and by the 

 fact that it explains what has hitherto been regarded as an in- 

 soluble problem. Distribution is an extremely slow process, 

 allowing time for acclimatisation, and the effect of all the various 

 factors that act upon it is to cause it to take place at a regular 

 rate, so that it becomes a measure of age, or vice versa. Barriers 

 alone interfere with it, but they may be of many kinds. 



Summary 

 The monotypic genera are very numerous, being 4853 out of 

 12,571 in the world, or 38-6 per cent. The ditypes are also nume- 

 rous, but are only 1632, or a drop of over 3000 from the mono- 

 types, while there is another drop to the 921 tritypes, and then 

 the numbers of genera of different sizes taper away in a long tail 

 to Astragalus at 1600 species. The mono- and di-types include 

 more than half of the total, and a very regular hollow curve is 

 formed. The individual families are arranged in the same way, 



13 



