Dec., 1923] CAMPBELL — AUSTRALASIAN BOTANICAL NOTES 535 
One of these, the Kermadec ridge, extends as far north as the Samoan group, 
while the other trends toward New Caledonia and tropical Australia. Lord 
Howe Island and Norfolk Island are connected with this ridge, and their 
flora reveals unmistakable relationships to that of New Zealand. 10 
Whether such truly Australian types as the Epacridaceae and Candollea- 
ceae reached New Zealand via this northern route, or whether they came 
from the south, we have no means of determining at present. 
How the conspicuous Fuegian flora reached Australasia has been the 
subject of much discussion, and there is difference of opinion on this subject. 
To the writer it seems inconceivable that the majority of the many common 
types could have been transported except by means of more or less con¬ 
tinuous land connections. This conclusion is emphasized by the numerous 
correspondences in the fauna, especially such forms as fresh-water fishes 
and Crustacea, which could hardly have migrated across several thousand 
miles of open ocean. 11 
The simplest explanation of the correspondence in both the fauna and 
flora of these widely separated regions would be the assumption of a former 
northward extension of the present Antarctic continent, or perhaps the 
existence of an extensive archipelago occupying this region. This land 
might be supposed to have developed a more or less uniform flora, such as 
existed in similar latitudes of the northern hemisphere in the late Tertiary. 
The subantarctic genera, like Nothofagus, Laurelia, Gunnera, etc., may be 
the remnants of this ancient antarctic flora which have survived in their 
present widely sundered habitats. Possibly also the ancestors of other 
types now confined to Australia or New Zealand may have had a similar 
origin. 
This view is strengthened by the discovery of plant fossils in the existing 
antarctic regions. Especially interesting was the discovery by a Swedish 
expedition on Seymour Island, near the Antarctic Continent, of fossils closely 
related to living New Zealand species. Among these were remains of a 
Knightia, apparently very similar to K. excelsa —one of the two living New 
Zealand Proteaceae. 
A migration of these ancient antarctic plants into South America, Austra¬ 
lasia, and South Africa, and the subsequent complete isolation of the 
respective regions, would explain many of the apparent anomalies in the 
distribution of the existing subantarctic floras, as well as the source of the 
ancestors of the peculiar floras of Australia and South Africa. 12 
The student of the New Zealand flora is at once impressed by the exten¬ 
sion of so many tropical and subtropical genera far beyond the latitudes 
10 Oliver, W. R. B. The vegetation and flora of Lord Howe Island. Trans. N. Z. 
Inst. 49: 24-161. 1915. 
11 Chilton, C. The subantarctic islands of New Zealand. 2 vols. Wellington, 1909. 
12 Dr. J. C. Willis has published a series of papers (Annals of Botany, 1916-1920) in 
which he discusses in great detail the flora of New Zealand and the sources from which it has 
been derived. His conclusions agree closely with the views expressed in the present paper. 
35 
