110 THE FURTHER EXTENSION OF THE [pt. ir 



sort into sizes the genera confined to any section of the world, 

 whether it be an individual island, or a larger area of territory 

 like Africa or South America, or whether it be the entire world 

 itself. Always there are many monotypes with a rapid drop 

 through the ditypes and tritypes, and a longer or shorter tail of 

 larger genera. 



The supposition that endemic genera are usually rehcs, as well 

 as the other that they are usually local adaptations, must be 

 ruled out of consideration in view of the facts brought up, and 

 the only supposition that at present seems at all feasible is that 

 provided by Age and Area, that in general they are young 

 beginners. This is also shown by the fact that the proportions 

 upon islands in the different families are not unlike the pro- 

 portionate sizes of these families in the world. 



Passing on to Monotypic Genera in Chapter xvii, it is shown 

 that these, which are usually much localised, display the same 

 phenomena. They are very numerous, over 38 per cent, of the 

 genera of the world containing only one species each, while there 

 are about 13 per cent, of ditypes, these two therefore containing 

 more than half the genera in the world. The proportion of mono- 

 types falls off with increasing size of area, and the proportions 

 oKgenera of other sizes bear a definite relation to that of mono- 

 types, showing that to explain these in general as reUcs or as 

 special adaptations would be absurd. They must usually be 

 voung beginners. 



Not only do these numbers, when plotted, exhibit a beautiful 

 hollow cun-e for the distribution into sizes of the genera of the 

 world, but the same thing is shown by every individual family. 

 Other arithmetical relationships between the monotypes and 

 other genera, depending upon the size of the area considered, 

 are also pointed out. 



Chapter xviii deals with the Hollow Curve of Distribution 

 and shows, by summing up what has already been said, how 

 universal this type of curve is, not only in the distribution of 

 species and genera (endemic or not) by area— Geographical Dis- 

 tribution or Distribution in Space— but in the distribution of 

 genera into groups according to their number of species— Evo- 

 lution or Distribution in Time. It is clearly evident throughout, 

 and usually in a very marked and unmistakable way, and goes 

 to show that Evolution and Geographical Distribution have gone 

 on "mechanically." The former appears to have been organised 

 at the start upon a definite plan, and its further unfolding, and 



