28 o 
Willis .- — The Floras of the Outlying Islands of 
There remains one species in Class 9, Ipomoea palmata , which ranges 
about 80 miles from North Cape. The only other species of Ipomoea in 
this region, /. biloba , occurs only in the Kermadecs. It is fairly clear from 
the very short range that I. palmata probably did not reach New Zealand 
from the Kermadecs by land. It is more probable that it has only com- 
paratively lately, so to speak, arrived, and by water carriage. There are 
hundreds of miles of the northern coast on which it would in all probability 
be equally at home. 
It is thus fairly evident that while a large part of the flora of the 
Kermadecs was in all probability derived from New Zealand, it is at least 
extremely probable that some of it was derived directly from some part of 
Polynesia, probably by way of the Tonga-Fiji ridge. This supposition is 
confirmed by the behaviour of the ferns, already pointed out ( 9 . p. 341). A 
number of Polynesian ferns appear to have entered New Zealand by way of 
the Kermadecs, and in New Zealand only range part of the length of the 
islands. 
When we come to compare the Kermadec flora with that comprised in 
the northern invasion (10, p. 356), and find that 22 of the 33 families are 
missing, it seems hardly possible, in view of what we now know about the 
rarity of dying out among flowering plants at any rate, that the northern 
invasion passed by way of the Kermadecs. On the other hand, we have 
seen that the only members of Anacardiaceae, Cucurbitaceae, and Pipera- 
ceae, three of the 33 families enumerated in it, probably passed to New 
Zealand by way of the Kermadecs. It would seem therefore probable that 
these three families should be excluded from the list of the northern 
invasion proper, and should be classed, together with the other plants 
mentioned above {Siege sbeckia, &c.), as a second northern (Kermadec) 
invasion. 
From the extreme regularity of the distribution of the Kermadec 
plants in New Zealand (except the conspicuously different Ipomoea palmata ) , 
one is obliged, it seems to me, to conclude that there was at one time a land 
bridge connecting the two. 
We also know (10, p. 358) that the northern invasion was mainly 
(84 per cent.) trees and shrubs. Examination of the Kermadec flora, 
however, shows that it contains 45 herbs and only 26 shrubs and trees. 
The latter thus form only 36 per cent, of the flora, a very different propor- 
tion from that in the northern invasion proper. 
Of the 2 6 shrubs and trees, 12 are confined to the Kermadecs, or at 
least do not reach New Zealand, and 5 of the other 14 range from end to 
end of New Zealand, the mean range of the whole 14 being represented by 
the figure 3*0. Of the 45 herbs, 7 only are not found in New Zealand, 
20 range New Zealand from end to end, and the mean range is represented 
by the figure 2-2. These facts are inclined to suggest, though they offer no 
