Stewart Island (. New. Zealand ). 483 
in the light thrown upon geographical distribution by age and area can no 
longer be accepted as sufficient to account for such phenomena. 
Just as we found the flora of Stewart, as older, to be composed of the 
larger (in general, older) families and genera of New Zealand proper, so here 
we shall expect the flora of these islands to be composed of the larger families 
and genera of the Stewart flora, and that of the Solanders especially so. 
Testing this we find that of the New Zealand families 22 are above the 
average in size (in New Zealand) and 69 below. Of the former 21 (95 per 
cent.) occur in Stewart, of the latter only 39 (56 per cent.) Of the Stewart 
families 15 are above the average (in Stewart) and 45 below. Of the former 
1 3 (86 per cent.) occur in the islets, of the latter 17 (37 per cent.). Of the 
islet families 9 are above the average and 21 below. Of the former 6 (66 
per cent.) occur in the Solanders and of the latter 4 (19 per cent.) only. It 
is thus clear that on the average a family is represented everywhere in pro- 
portion to its size in the neighbouring country. We may put this in another 
way, thus. The average size in New Zealand and the surrounding islands of 
a family occurring there (91 fams., 1,392 species) is 15 species. The average 
size in New Zealand of families occurring in Stewart is 21 species, or much 
higher. The average size in New Zealand of the families that occur in the 
islets now under consideration is 35 species (30 fams., 1,059 species). And 
finally, the average size in New Zealand of the families occurring in the 
Solanders (11 fams., 765 species) is 69 species. The figures thus form 
a progressive series, showing clearly that on the whole the larger in New 
Zealand a family is, the greater in the New Zealand area is its r&nge. 
We shall further expect that, as usual, there will be more families in 
proportion to genera, and more genera in proportion to species, the farther 
out we go from the centre of New Zealand. 
Table II. 
Fams. 
Gen. 
Spp. 
Gen. per fam. 
Spp. per 
New Zealand 
9 1 
329 
1392 
3*6 
4 • 2 
Stewart 
60 
l6y 
3 8 3 
2.8 
2.2 
Islets 
30 
56 
98 
1.8 
i *7 
Solanders 
11 
13 
15 
1. 1 
1. 1 
The prediction is fully borne out. 
As one goes outward from New Zealand in this way, the plants will on 
the average become steadily older, so that one will expect to find the pro- 
portion in common with the outlying islands (Kermadecs, Chathams, Auck- 
lands), which also have old floras, steadily increasing. Testing this gives 
Table III. 
Occur in Reach K., Ch ., or Au. %. 
New Zealand, 1301 
199 
* 5 
Stewart, 383 
Jr 3 
40 
53 
Islets, 98 
Solanders, 1 5 
52 
9 
60 
