has been brought about rather by extermination than specialization,
and their distinctive facies by the development and multiplication
of the surviving types.
The distribution of mountain barriers in the Old and New Worlds is in striking contrast. In the former they run from east to west; in the latter from north to south. In the Old World the boreal zone is almost sharply cut off and afforded no means of escape for the Miocene vegetation when the climate became more severe. Thus in the Mediterranean region the large groups of palms, figs, myrtles and laurels are each only represented by single surviving species. The great tropical family of the Gesneraceae has left behind a few outliers: Ramondia in the Pyrenees, Haberlea in the Balkans, and Jankaea in Thessaly; the Pyrenees also possess a minute Dioscorea, sole European survivor of the yams of the tropics.
In North America there is no such barrier, the Miocene flora has been able to escape by migration the fluctuations of climate and to return when they ameliorated. It has preserved its characteristic types, such as Magnolia, Liriodendron, Liquidambar, Torreya, Taxodium and Sequoia. While it has been customary to describe the Miocene flora of Europe as of a North American type, it would be more accurate to describe the latter as having in great measure preserved its Miocene character.
If mountains serve as barriers which arrest the migration of the vegetation at their base, their upper levels and summits afford lines of communication by which the floras of colder regions in the northern hemisphere can obtain a southern extension even across the tropics. They doubtless equally supply a path by which southern temperate types may have extended northwards. Thus the characteristic assemblage of plants to which Sir Joseph Hooker has given the name Scandinavian “is present in every latitude of the globe, and is the only one that is so” (Trans. Linn. Soc. xxiii. 253). In the mountains of Peru we find such characteristic northern genera as Draba, Alchemilla, Saxifraga, Valeriana, Gentiana and Bartsia. High elevations reproduce the physical conditions of high latitudes. The aqueous vapour in the atmosphere is transparent to luminous but opaque to obscure heat-rays. The latter are retained to warm the air at lower levels, while it remains cold at higher. It results that besides a horizontal distribution of plants, there is also an altitudinal: a fact of cardinal importance, the first observation of which has been attributed to Tournefort.
Speaking generally, all plants tend to exhaust particular constituents of the soil on which they grow. Nature therefore has provided various contrivances by which their seeds are disseminated beyond the actual position they occupy. In a large number of cases these only provide for migration within sufficient but narrow limits; such plants would be content to remain local. But other physical agencies come into play which may be briefly noticed. The first of these is wind: it cannot be doubted that small seeds can be swept up like dust and transported to considerable distances. This is certainly the case with fern-spores. The vegetation of Krakatoa was completely exterminated in 1883 by a thick coat of red-hot pumice. Yet in 1886 Treub found that it was beginning to cover itself again with plants, including eleven species of ferns; but the nearest source of supply was 10 m. distant. Seeds are carried with more facility when provided with plumes or wings. Treub found on Krakatoa four species of composites and two grasses. Water is another obvious means of transport. The littoral vegetation of coral islands is derived from sea-borne fruits. The seeds of West Indian plants are thrown on the western shores of the British Isles, and as they are capable of germination, the species are only prevented from establishing themselves by an uncongenial climate. Travers picked up a seed of Edwardsia in the Chatham Islands, evidently washed ashore from New Zealand (Linn. Soc. Journ. ix. 1865). Rivers bring down the plants of the upper levels of their basins to the lower: thus species characteristic of the chalk are found on the banks of the Thames near London. Birds are even more effective than wind in transporting seeds to long distances. Seeds are carried in soil adhering to their feet and plumage, and aquatic plants have in consequence for the most part an exceptionally wide range. Fruit-pigeons are an effective means of transport in the tropics by the undigested seeds which they void in their excrement. Quadrupeds also play their part by carrying seeds or fruits entangled in their coats. Xanthium spinosum has spread from the Russian steppes to every stock-raising country in the world, and in some cases has made the industry impossible. Even insects, as in the case of South African locusts, have been found to be a means of distributing seeds.
The primary regions of vegetation, already indicated, and their subordinate provinces may now be considered more in detail.
I. North Temperate Region.—Many writers on the distribution of animals prefer to separate this into two regions of “primary rank”: the Palaearctic and the Nearctic. But to justify such a division it is necessary to establish either an exclusive possession or a marked predominance of types in the one which are correspondingly deficient in the other. This cannot apparently be done for insects or for birds; Newton accordingly unites the two into the Holarctic region. It equally fails for plants. To take, for example, one of the most characteristic features of the Palaearctic region, its catkin-bearing deciduous trees: in North America we find precisely the same genera as in the Old World—oaks, chestnuts, beeches, hazels, horn beams, birches, alders, willows and poplars. Or to take the small but well defined group of five-leaved pines, all the species of which may be seen growing side by side at Kew under identical conditions: we have the Weymouth pine (Pinus Strobus) in eastern North America, P. monticola and the sugar pine (P. Lambertiana) in California, P. Ayacahuste in Mexico, the Arolla pine (P. Cembra) in Switzerland and Siberia, P. Peuce in Greece, the Bhotan pine (P. excelsa) in the Himalayas, and two other species in Japan. Amongst broad-leaved trees Juglans has a similar Holarctic range, descending to the West Indies; so has Aesculus, were it not lacking in Europe; it becomes tropical in South America and Malaya. If we turn to herbaceous plants, Hemsley has pointed out that of the thirteen genera of Ranunculaceae in California, eleven are British.
While the tropics preserve for us what remains of the pre-Tertiary or, at the latest, Eocene vegetation of the earth, which formerly had a much wider extension, the flora of the North Temperate region is often described as the survival of the Miocene. Engler therefore calls it Arcto-Tertiary. We must, however, agree with Starkie Gardner that it is only Miocene as regards its present position, which was originally farther north, and that its actual origin was much earlier. There has been in effect a successive shifting of zones of vegetation southwards from the pole. Their distinctive and adaptive characteristics doubtless began to be established as soon as the phanerogamic flora was constituted. There is no reason to suppose that the peculiarities of the arctic flora are more modern than those of any other, though there is no fossil evidence to prove that it was not so.
The North Temperate region admits of subdivision into several well-marked sub-regions. The general method by which this is effected in this and other cases is statistical. As A. de Candolle, however, points out, exclusive reliance on this may be misleading unless we also take into account the character and affinities of the plants dealt with (Geogr. Bot. i. 1166). The numerical predominance of certain families or their absence affords criteria for marking out boundary lines and tracing relationships. The analysis of the flora of the British Isles will afford an illustration. This was first attempted in 1835 by H. C. Watson, and his conclusions were enforced ten years later by Edward Forbes, who dealt also with the fauna. Watson showed that Scotland primarily, and to a less extent the north of England, possessed species which do not reach the south. Such are the crow berry (Empetrum nigrum), Trientalis europaea, Rubus saxatilis and the globe-flower (Trollius europaeus). He further found that there was an element which he termed “boreal . . . in a more intense degree,” which amounted to about “a fifteenth of the whole flora.” This was not confined to the north but may occur on the mountains of England and Wales: Salix herbacea, Silene acaulis and Dryas octopetala will serve as examples. Even so small an area as that of Britain illustrates what has already been pointed out, that the species of a flora change both with latitude and altitude. Watson further brought out the striking fact that the west and east of Britain each had species peculiar to it; the former he characterized as Atlantic, the latter as Germanic. The Cornish heath (Erica vagans) and the maiden-hair fern (Adiantum Capillus-Veneris) may serve as instances of the one, the man-orchis (Aceras anthropophora) and Reseda lutea of the other. Ireland illustrates the same fact. It possesses about 1000 species, or about two-thirds the number of Britain. On its western shores there are some twenty, such as Saxifraga umbrosa, Erica mediterranea and Arbutus unedo, which are not found in Britain at all. The British Phanerogamic flora, it may be remarked, does not contain a single endemic species, and 38% of the total number are common to the three northern continents.
The analysis of larger areas yields results of the same kind. Within the same region we may expect to find considerable differences as we pass from one meridian to another. Assuming that in its circumpolar origin the North Temperate flora was fairly homogeneous, it would meet in its centrifugal extension with a wide range of local conditions; these would favour the preservation of numerous species in some genera, their greater or less elimination in others. Thus comparing the Nearctic and Palaearctic floras we find striking differences overlying the points of agreement already indicated. The former is poor in Cruciferae, Caryophyllaceae, Umbelliferae, Primulaceae and Labiatae; but for the occurrence of Calluna, in Newfoundland it would have no heaths. On the other hand, it is rich in Compositae, especially Solidago and Aster, Polemoniaceae, Asclepiadaceae, Hydrophyllaceae and Cyperaceae, and it has the endemic Sarracenia, type of a family structurally allied to poppies, of which of the remaining genera Darlingtonia is Californian, and Heliamphora Venezuelan. These distinctions led Sir Joseph Hooker to claim for the two divisions the rank of primary regions. Darwin doubted, however, whether they ought to be separated (Life, iii. 230). Lyell, discussing the facts of zoological distribution, admits that “the farther we go north . . . the more the discordance in genera and species diminishes” (Principles, ii. 340); and Hemsley finds that not less than 75% of the genera in the flora of eastern North America “are represented in the old world,” and, according to Asa Gray, 50% in Europe.
Latitudinally the region subdivides naturally into several well marked sub-regions which must be briefly discussed.
