New Zealand and their Distribution. 
287 
In Campbell, the Antipodes, and Macquarie we find in general the 
Auckland flora, in less and less proportion, so that on the whole these 
islands must probably have received their floras by the same invasion. It 
would lead too far to go into greater detail. 
The Chatham Islands. 
To turn lastly to the Chathams, it is obvious that they could never 
have formed part of a bridge from New Zealand to anywhere else, the .water 
on their eastern side being almost the deepest in the world, so that their 
flora, except in so far as it may have been brought by currents, must have 
come through New Zealand. Even if the whole area shown in the map as 
above the depth of 1,000 fathoms were dry land, a species beginning 
between the Chathams and Antipodes would as a rule have also reached 
New Zealand by the time that it had reached both groups of islands. 
The Chathams are fairly opposite to the middle of New Zealand, and 
connected by shallower water with the South than with the North Island, so 
that one will expect their flora to be fairly rich, unless they were cut off 
very early indeed, a matter as to which we have no information. As 
a matter of fact they have the richest flora of the outlying island groups, 
composed of 106 species of Dicotyledons and 49 of Monocotyledons, the 
proportion of the latter being, as we have seen, intermediate between that 
in the Kermadecs and that in the Aucklands. 
(29) We shall expect that on the whole the families missing in the 
Chathams from the Stewart list, which is likewise old, will be the smaller 
families of that island, i. e. on the whole those which have been the latest in 
arriving there. If we omit all those with one species, we omit Magnoliaceae, 
Pittosporaceae, Elatinaceae, Hypericaceae, Linaceae, Cornaceae, Goodenia- 
ceae, Primulaceae, Apocynaceae, Loganiaceae, Labiatae, Illecebraceae, 
Lentibulariaceae, Chloranthaceae, Loranthaceae, and Euphorbiaceae. 
Primulaceae must certainly occur in the Chathams, as the only species in 
the family occurs in both Kermadecs and Aucklands. Leaving this out, 
the only other families in this list which are recorded from the Chathams 
are Linaceae, Cornaceae, Labiatae, and Euphorbiaceae. The other Stewart 
families not recorded from the Chathams are Portulacaceae (2 species), 
Tiliaceae (4), Saxifragaceae (3), Droseraceae (4), Stylidiaceae (4), Erica- 
ceae (2), Plantaginaceae (3), and Centrolepidaceae (2), but we have already 
seen that (prediction 5) we must expect the Chathams not to contain the very 
small families of the southern invasion. This excludes from this list Portu- 
lacaceae, Droseraceae, Stylidiaceae, Plantaginaceae, and Centrolepidaceae, 
leaving only three families of the Stewart flora that might be expected and 
have not been'recorded, and four recorded that would hardly be expected, 
as being families with only one species in Stewart. The seven families that 
occur in the Chathams and do not as yet appear to have been recorded 
