290 DISTRIBUTION [ VEGETABLE. have become geograpliical) ; and, as mentioned above, there are a few scattered species, representing the South African flora, iu extra-tropical South America. There are, however, two offshoots in a northern direction. The remarkable West European flora, already referred to, pos sesses species of Erica, shrubby Legummosce, Lobelia, Gla diolus, &c., " more nearly added to corresponding Cape species than they are to each other." The other extension is to Eastern Africa. The sub-alpine vegetation of Iviliuia Njaro is distinctly South African, and Hooker suggests " the probability of the South African flora being repre sented all along the highlands of Eastern Africa, from Natal to Abyssinia ; and further, seeing that most of- the South African plants found in the Cameroons are also natives of Abyssinia, it would appear probable that the migration of these to the Cameroons was by and through Abyssinia." l The further suggestion that this may have been the path travelled by the West European extension of the South African flora is sufficiently obvious. The amount of agreement amongst these scattered frag ments of a great flora points necessarily to a state of things when the lands they now occupy were at one time or other in more or less of intimate connection. The amount of differentiation between the floras, and the fact that agree ment has to be sought in groups of high rather than of small rank, points equally to the fact that such connections must have been far from recent. 2 The detailed study of separate groups leads by another path to the same result, and, as a good instance of the new phase into which taxo- nomic botany is entering in the light of the study of geographical distribution, reference may be particularly made to Bantham s important investigation into the past history and migrations of the Campamdacece. 3 III. THE TROPICAL FLORA. This is still perhaps too imperfectly known to admit of any very plausible general ization. It obviously presents three great subdivisions. 1. The Indo-Malayan extends from the Himalayas to north-east Australia and Japan. In the latter country it meets the northern temperate flora, from which in India it is sharply divided by the Himalayas. 2. The American is still a perfect mine of unexplored botanical wealth. Beutham remarks " No general com parison of Asiatic and American tropical vegetation can therefore be made without immense labour of detail. As far as we know, however, the resemblance between them is only in some of the races of a higher grade, natural orders and comprehensive genera ; the smaller genera and species, and many even of the higher ones, are totally different ; or if a few species are identical, they are gene rally, if woody or arborescent such as Entada, Gyrocarpus, <fcc., wholly or partially maritime, and may have traversed the ocean during its present configuration, or if herba ceous widely, spread weeds still more likely to be spread all round the tropics under existing conditions." 4 There are, however, some extraordinary points of connection between the tropical floras of the Old and New Worlds, to which there is at present scarcely any clue. Thus Ternstroemia emarginata, endemic to Ceylon, so closely resembles the Brazilian Ternstroemia cuneifolia as to be barely distinguishable. 3. The African tropical flora is probably the most 1 Cf. also Darwin, Origin of Species, 4th ed. p. 474. 9 Journ. Linn. Soc. Lot., xiv. p. 145. Journ, Linn. Soc* Hot., xv. p. 11. 4 Presidential address, 1869, p. 24. imperfectly known of any. Bentham considers it as uf great antiquity, and as having preserved large numbers of persistent types from which races " have widely diverged in Asia or America, or in both." He further remarks that "as our knowledge of the vegetation of tropical Africa has increased, we have discovered a greater number of Asiatic types; but still there are, even in the interior, a certain number of American ones, offering a problem the solution of which has scarcely been attempted." 5 In Composite? American genera are represented in east tropical Africa, and Benthara is led on various grounds to regard this as the principal area of preservation of the most ancient tropical flora of the Old World. 6 A well-marked eastern element in the African tropical flora is generally accepted. Madagascar, whose flora bears the marks of long isolation, contains Malayan and even Australian types; and it is a problem worth future inquiry whether the connection between the floras of tropical America and Africa may not have taken place south of the tropics, and by similar (though more northern) paths to those which once united the scattered members of the great Southern flora, As might have been expected, during the Tertiary period the tropical flora extended much beyond its present limits. De Saporta, who has studied with great caution the fossil floraof the gypseous beds(Eocene)of Aix in Provence, arrives at the following conclusions : 7 The principal families were such as characterize tropical vegetation, especially Indian - Ebenaceve, Anacardiacece, Sapindacece, Sterculiacece, Let/u- minosoe. The affinities of the ancient vegetation of Aix in respect of generic types, general facies, and composition with that of India and the Indian archipelago, Ciuna, the Philippines, and Japan at the present day, are in perfect accordance with the theory that these regions formed the shores of our ancient nummulitic sea, extending from Morocco to Japan, and entirely comprised in the tropical zone of the Eocene world, which extended to the 55th parallel. Besides its relation to South-Eastern Asia, the Aix flora exhibits, according to De Saporta, a strong affinity with that of Africa, lying between Abyssinia and the Cape, of which, however, it must be confessed, but little is as yet known. Here this outline of the present state of a most impor tant and rapidly developing branch of biological science must be concluded. The writer has availed himself very freely of the kind permission of Mr Bentham perhaps the greatest living master of the subject to make use of his scattered but invaluable papers, not scrupling to borrow from them all that seemed most important and sugges tive, but has generally thought it fairer both to the subject and to Mr Bentham to do so in his own words. For two heads of the subject it must suffice merely to give refer ences. On the remarkable phenomena of insular floras, the reader should consult Sir Joseph Hooker s well-known lecture delivered before the British Association in 1866, and printed in the Gardener s Chronicle for January 18G7, or, in default of this, the summary given in LyelPs Prin ciples of Geology, 10th ed., vol. ii. pp. 417-421. On the means of dispersion of plants, reference may also be made to LyelPs work already quoted, vol. ii. pp. 386-400 ; Dar win s Origin of Species, 4th ed., pp. 425-442 ; Bentham, Presidential Address, 1869, pp. 7, 8. (w. T. T. D.) B Journ. Linn. Soc. Bot., xiii. 545. 6 Uentham, I. c. p. 24.
7 Annales des Sciences Naturelles, Sept. 1, 1872.