180 ENDEMISM AND DISTRIBUTION: GENERA [pt. ii 



Africa, and South America, and (4) all the genera of the world. 

 Arranging the families in groups of ten in the order of their size 

 in the world (as judged by number of genera), and taking for 

 each of the other three areas the number of genera in the sa?ne 

 ten families, one gets the following table : 



Table showing in each pair of columns the numher, and the per- 

 centage, of genera that occur in the world, and thai are confined 

 to three sections of it {ending with those confined to the islafids). 

 The first horizotital line shows the figures for the ten largest 

 families in the world for each of these, and the following lines 

 those for the second, third, etc. tens of families in the xoorld. The 

 percentages are counted downwards ; the 40 per cent, at the top 

 of the first column means 40 per cent, of the genera of the world. 



The percentages agree with one another in the four columns 

 in the most remarkably close manner, as a little inspection will 

 soon show. The greatest difference in the whole table occurs in 

 the second line, between 14-6 per cent, for Australia, etc., and 

 18-0 per cent, for islands, a difference of 3-4 per cent. The second 

 greatest is in the first line, between 38-3 for islands and 40-5 for 

 West Australia, etc., a difference of 2-2 only. 



If these percentages be plotted as curves, they give the re- 

 markable figure shown. 



The close coincidence of these (hollow) curves is very remark- 



