New Zealand and their Distribution . 275 
much space to give the whole list over again (cf. 11, p. 27), but of the 382 
species so far recorded from Stewart, no fewer than 153, or 40 per cent., are 
found in the islands, whereas of the 1,301 species found in New Zealand as 
a whole, only 199 occur on the islands, or 15 per cent. Thus there are only 
46 species bn the islands, other than local endemics or species which do not 
occur at all in New Zealand, which do not occur in Stewart, against 153 
which do. 
(13) We may predict that the families which occur in Stewart but 
have no species on the islands will on the whole be small (in Stewart). 
The average size of a family in Stewart is 6*3 species, and those families 
that contain no species on the islands contain in Stewart 4, 3, 2, 1, 1, 1, i, 1, 
1, 1, 1 species, or all below the average size in Stewart. 
(14) We may also predict in this connexion that the islands will have 
very many species in common between themselves, and this also proves to 
be the case. The Kermadecs have a total flora of 71, of which 19 occur 
nowhere else, leaving 52, of which 30 occur in the Chathams and 5 in the 
Aucklands. The Chathams have a total flora of 152, of which 27 occur 
nowhere else, leaving 125, of which 30 occur in the Kermadecs, and 32 in 
the Aucklands. The Aucklands have a total flora of 119, of which 44 do 
not occur in New Zealand, leaving 75, of which 32 occur in the Chathams 
and 5 in the Kermadecs. 
(15) We may predict that the species in common between the islands, 
or between the islands and Stewart, will show a greater proportion of wides 
(which on the whole are older than endemics), the more the islands on which 
the species occur. 
Table X. 
Wides. 
Proportion. 
Endemics. 
Proportion. 
3 islands and Stewart 
4 
80% 
1 
20 / 
2 >> » 
27 
64% 
55 
36 % 
1 >5 J 5 >> 
43 
4 i % 
61 
59 % 
No „ but „ 
57 
2 4 % 
174 
76 % 
Not even reaching Stewart 
170 
18% 
749 
82 % 
The proportions decrease and increase, respectively, with regularity. 
(16) We may now go on to some slightly more general predictions, 
and say that as on the whole the islands were probably cut off earlier than 
Stewart, their plants will on the whole be older, and should therefore show 
a greater proportion of families to genera and genera to species. 
Table 
XI. 
Families. 
Genera. 
Gen. per fam. 
Species. 
Spp. per gen. 
Islands 
55 
130 
2 -3 
199 
i *5 
Stewart 
60 
169 
2.8 
382 
2.2 
New Zealand 
9 i 
3 2 9 
3-6 
1392 
4.2 
(17) We may make the same prediction with regard to the islands as 
the second made with respect to Stewart on p. 34 of 11 , that on the whole 
