species are known out of a probable total of not less than 12,000,
and of these more than half are endemic. The number of species
to a genus, 3, is only half that found in other large areas. This
aggregation of genera and of endemic species is characteristic of the
circumferential portions of the earth’s land surface, the expansion
of the one and the further advance of the other is arrested. The
sub-region is probably sharply cut off from the Intermediate.
Maximowicz finds that 40% of the plants of Manchuria are common
to Europe and Asia, but the proportion falls sharply to 16% in the
case of Japan. Its connexion with the Mediterraneo-Oriental
sub-region is still more remote. China has no Cistus or heath, only a
single Ferula, while Astragalus is reduced to 35 species. There are
to species of Pistacia and four of Liquidambar. The affinity to
Atlantic North America is strongly marked as it has long been known
to be in Japan. China has 66 species of Quercus, 35 of Vitis, 2 of
Aesculus, 42 of Acer, 33 Magnoliaceae (including two species of
Liriodendron), 12 Anonaceae, 71 Ternstroemiaceae (including the
tea-plant), and 4 of Clethra, which has a solitary western representative
in Madeira. Trachycarpus and Rhapis are characteristic
palms, and Cycadeae are represented by Cycas.
5. The Mexico-American sub-region has as its northern boundary the parallel of lat. 36° as far as New Mexico and then northwards to the Pacific coast at lat. 40°. The eastern and western halves are contrasted in climate—the former being moist and the latter dry—and have been distinguished by some zoologists as distinct sub-regions. They are in fact in some degree comparable to sub-regions 3 and 4 in the Old World. The absence of marked natural boundaries makes any precise north and south limitation difficult. But it has been a centre of preservation of the Taxodieae, a tribe of Coniferae of great antiquity. Taxodium (with single species in China and Mexico) is represented by the deciduous cypress (T. distichum), which extends from Flor1da to Texas. The two species of Sequoia, the “Wellingtonia” (S. gigantea) and the redwood (S. sempervirens), are confined to California. In the eastern forests the prevalence of Magnoliaceae and of Clethra and Rhododendron continues the alliance with eastern Asia. Florida derives a tropical element from the Antilles. Amongst palms the Corypheae are represented by Sabal and Thrinax, and there is a solitary Zamia amongst Cycads. The western dry areas have the old-world leguminous Astragalus and Prosopis (Mesquit), but are especially characterized by the northward extension of the new-world tropical Cactaceae, Mammillaria, Cereus and Opuntia, by succulent Amaryllideae such as Agave (of which the so-called “American aloe” is a type), and by arborescent Liliaceae (Yucca). Amongst palms Washingtonia, Brahea and Erythea (all Corypheae) replace the eastern genera. On the west coast Cupressus Lawsoniana replaces the northern Thuya gigantea, and a laurel (Umbellularia of isolated affinity) forms forests. California and Arizona have each a species of Platanus, a dying-out genus. Elsewhere it is only represented by P. occidentalis, the largest tree of the Atlantic forests from Maine to Oregon, and by P. orientalis in the eastern Mediterranean. Otherwise the Californian flora is entirely deficient in the characteristic features of that of eastern North America. Nor, with perhaps the interesting exception of Castanopsis chrysophylla, the solitary representative in the New World of an east Asiatic genus, which ranges from Oregon to California, has it any affinity with the Chino-Japanese sub-region. Its closest connexion is with the South American Andine.
II. The Tropical Region.—The permanence of continents and great oceans as first insisted upon by J. D. Dana, but, as already stated, has later received support on purely physical grounds. It precludes the explanation of any common features in the dissevered portions of the tropical area of vegetation by lateral communications, and throws back their origin to the remotest geological antiquity. In point of fact, resemblance is in the main confined to the higher groups, such as families and large genera; the smaller genera and species are entirely different. No genus or species of palm, for example, is common to the Old and New Worlds. The ancient broad-leaved Gymnosperm Gnetum has a few surviving species scattered through the tropics of both worlds, one reaching Polynesia.
1. African sub-region.—Western Arabia must be added to the African continent, which, with this exception and possibly a former European connexion in the far west, has had apparently from a very early period an almost insular character. Bentham remarks (Journ. Linn. Soc. xiii. 492): “Here, more perhaps than in any other part of the globe, in Compositae as in so many other orders, we may fancy we see the scattered remains of ancient races dwindling down to their last representatives.” It is remarkable that the characteristic features of the Miocene flora, which in other parts of the world have spread and developed southwards, are conspicuously absent from the African tropical flora. It has no Magnoliaceae, no maples, Pomaceae, or Vacciniaceae, no Rhododendron and no Abietineae. Perhaps even more striking is the absence of Cupuliferae; Quercus, in particular, which from Tertiary times has been a conspicuous northern type and in Malayan tropical conditions has developed others which are widely divergent. Palms are strikingly deficient: there are only three out of 79 genera of Areceae, and the Corypheae are entirely absent. But including the Mascarene Islands and Seychelles the Borasseae are exclusively African. Aroideae are poorly represented compared with either South America or Malays. A peculiar feature in which tropical Africa stands alone is that at least one-fifth and probably more of the species are common to both sides of the continent and presumably stretch right across it. An Indian element derived from the north-east is most marked on the eastern side: the Himalayan Gloriosa will suffice as an example, and of more tropical types Phoenix and Calamus amongst palms. The forest flora of Madagascar, though including an endemic family Chlaenaceae, is essentially tropical African, and the upland flora south temperate.
2. The Indo-Malayan sub-region includes the Indian and Malayan peninsulas, Cochin-China and southern China, the Malayan archipelago, and Philippines, with New Guinea and Polynesia, excluding the Sandwich Islands. Probably in point of number of species the preponderant family is Orchideae, though, as Hemsley remarks, they do not “give character to the scenery, or constitute the bulk of the vegetation.” In Malaya and eastward the forests are rich in arborescent figs, laurels, myrtles, nutmegs, oaks and bamboos. Dipterocarpeae and Nepenthaceae only extend with a few outliers into the African sub-region. Screw pines have a closer connexion. Compositae are deficient. Amongst palms Areceae and Calameae are preponderant. Cycads are represented by Cycas itself, which in several species ranges from southern India to Polynesia. In India proper, with a dryer climate, grasses and Leguminosae take the lead in the number of species. But it has few distinctive botanical features. In the north-west it meets the Mediterraneo-Oriental and in the north-east the Chino-Japanese sub-regions, while south India and Ceylon have received a Malayan contribution. Bengal has no Cycas, oaks or nutmegs. Apart from the occurrence of Cycas, the Asiatic character of the Polynesian flora is illustrated by the distribution of Meliaceae. C. de Candolle finds that with one exception the species belong to genera represented in one or other of the Indian peninsulas.
3. The South American sub-region is perhaps richer in peculiar and distinctive types than either of the preceding. As in the Indo-Malayan sub-region, epiphytic orchids are probably most numerous in point of species, but the genera and even sub-tribes are far more restricted in their range than in the Old World; 4 sub-tribes with 74 genera of Vandeae are confined to South America, though varying in range of climate and altitude. Amongst arboreous families Leguminosae and Euphorbiaceae are prominent; Hevea belonging to the latter is widely distributed in various species in the Amazon basin, and yields Para and other kinds of rubber. Amongst Rubiaceae, Cinchoneae with some outliers in the Old World have their headquarters at cooler levels. In Brazil the myrtles are represented by “monkey-pots” (Lecythideae). Nearly related to myrtles are Melastomaceae which, poorly represented in the Old World, have attained here so prodigious a development in genera and species, that Ball looks upon it as the seat of origin of the family. Amongst Ternstroemiaceae, the singular Marcgravieae are endemic. So also are the Vochysiaceae allied to the “milkworts.” Cactaceae are widely spread and both northwards and southwards extend into temperate regions. Screw pines are replaced by the nearly allied Cyclanthaceae. The Amazon basin is the richest area in the world in palms, of which the Cocoineae are confined to South America, except the coco-nut, which has perhaps spread thence into Polynesia and eastward. The singular shrubby Amaryllids, Vellozieae, are common to tropical an South Africa, Madagascar and Brazil. Aroids, of which the tribes are not restricted in their distribution, have two large endemic genera, Philodendron and Anthurium. Amongst Cycads, Zamia is confined to the New World, and amongst Conifers, Araucaria, limited to the southern hemisphere, has scarcely less antiquity; Pinus reaches as far south as Cuba and Nicaragua.
The flora of the Hawaiian Islands has complicated relations. Out of the 860 indigenous plants, 80% are endemic, but Hillebrand finds that a large number are of American affinity.
III. The South Temperate Region contrasts remarkably with the northern. Instead of large continuous areas, in which local characteristics sometimes blend, it occupies widely dissevered territories in which specialization, intensified by long separation, has mostly effaced the possibility of comparing species and even genera, and compels us to seek for points of contact in groups of a higher order. The resemblances consist, in fact, not so much in the existence of one general facies running through the regions, as is the case with the northern flora, but in the presence of peculiar types, such as those belonging to the families Restiaceae, Proteaceae, Ericaceae, Mutisiaceae and Rutaceae.
1. The South African sub-region has a fora richer perhaps in number of species than any other; and these are often extremely local and restricted in area. It exhibits in a marked degree the density of species which, as already pointed out, is explicable by the arrest of further southern expansion. Hemsley remarks that “the northern genus Erica, which covers thousands of square miles in Europe with very few species, is represented by hundreds of species in a comparatively small area in South Africa.” There is an interesting connexion with Europe through the so-called Iberian flora. Bentham (Pres. addr. Linn. Soc., 1869, p. 25) points out that “the west-European species of Erica, Genisteae, Lobelia, Gladiolus, &c., are some of them more nearly allied to corresponding Cape species than they are to each other; and many of the somewhat higher races, groups of species and genera, have evidently diverged from stocks
