5°8 
Willis. — Endemic Genera of 
Fams. with endemic 
genera on islands. 
Fams. without 
With 1 
genus 3 
5 2 
genera 1 9 
49 
,, 0-10 
» 12 
24 
>> n-30 
» 36 
12 
„ 31-50 
,, 21 
2 
,, 51-80 
» !9 
2 
„ 81-150 
» 23 
r 
„ over 150 
» *7 
— 
which shows that the families with no island endemic genera are crowded 
towards the bottom of the list. Or yet again, one may take the total of 
genera in the families of the two lists, when one finds that the 150 families 
in the first list, that contain endemic genera on islands, contain in all 11,627 
genera, an average of 77 per family, while the 142 in the second list contain 
in all 890, or an average of 6 per family. 
It is quite clear that the facts do not bear out any prophecy based on 
the idea that endemics are moribund. The small families, and families of 
disconnected distribution, show no unusual prominence, but are on the 
contrary very badly represented. 
But the final proof against the moribund nature of endemic genera as 
a whole, both for islands and for other places in which they abound, such as 
South Africa or West Australia, is given by adding up all the genera found 
only in 
1. The islands of the World. 
2. West Australia, South Africa, and Brazil, all areas rich in endemic 
forms, and among them showing much variety in conditions, with forest, 
campo, and desert. 
3. Australia, Africa, and South America. 
4. The World. 
Just as in the case of the groups of islands, the individual families vary 
in their arrangement amongst themselves in the various lists ; indeed, were 
it not so, the fact would be perfectly astonishing. But it is generally found 
that any one family (unless herbaceous) is not very far apart in its situations 
in the different lists, and if we add the families together in groups of ten, local 
variation will be quite sufficiently covered. Taking as a standard the groups 
of ten into which the world list divides, and adding together in each of the 
other lists the genera found in the same ten families, we arrive at the table 
on p. 509. 
The figures in this table are very striking, and bear out in a very 
complete manner the prediction based upon the supposition that endemic 
genera cannot be survivals, except in rare cases. The percentage numbers 
are so close together in the different columns that the largest difference of 
all is that in the second line, between 18-0 for islands and 14*6 for Australia, 
Africa, and South America combined. It is clear in the first place that the 
islands, with few exceptions, and those almost certainly not the large 
