Feb. 1 1 , I 



NATURE 



339 



namely : — (i) Those with "vegetation comprising a large 

 endemic element, the nearest atSnities of which are not 

 always found in any one continent " ; to this category 

 belong St. Helena, Juan Fernandez, the Sandwich, 

 Galapagos, and Seychelles Groups. (2) Those with 

 " vegetation comprising a small, chiefly specific endemic 

 element, the origin of which is easily traced " ; here are 

 included the Bermudas, Azores, Ascension, the islands in 

 the southern part of the Indian Ocean, and the Admiralty 

 Islands. (3) Those with a "vegetation comprising no 

 endemic element" (which have become stocked with 

 plants in very recent times) ; such are the Keeling and 

 other coral islands in the Indian and Pacific Oceans A 

 statistical account of floras of several islands not treated 

 of in detail in the subsequent reports is given to illustrate 

 the feature characteristic of these several categories. 



A very prominent place is given by Mr. Hemsley to 

 the question of the "dispersal of plants by oceanic 

 currents and by birds." In the appendix to Part III. 

 there are descriptions of drift-seeds and seed-vessels, and 

 of seeds and fruits from the crops of birds, collected by 

 Mr. Moseley, Dr. Guppy, and others, to which is pre- 

 faced an historical resume of the subject. We have in 

 fact an epitome of all that has been written upon the 

 subject up to the present time. The evidence Mr. 

 Hemsley brings forward of the potency of these agencies 

 in plant-dispersion is irresistible, and eftectively over- 

 throws the opinion so frequently expressed by Alphonse 

 de Candoile — an opinion founded in great part upon the 

 capacity of seeds to retain vitality when immersed in 

 sea-water, as determined by experiment — that oceanic cur- 

 rents have played and play an unimportant part in plant- 

 diffusion. The views of the two botanists are placed in 

 striking contrast by a comparison of the list of species 

 certainly or probably dispersed by ocean currents given by 

 each : De CandoUe's contains about two dozen, Herasley's 

 over 100. Mr. Hemsley guards himself against being sup- 

 posed to regard the sea as the principal agent, or indeed 

 as anything more than a subordinate agent in bringing 

 about the present distribution of plants, " for it is mani- 

 fest that the action of currents and birds of passage 

 are insufficient to account for certain elements in 

 the vegetation of many islands." But at the same 

 time he goes so far as to maintain that the littoral 

 flora owes its present characteristics to the fact that 

 the seeds of the plants composing it are capable of 

 withstanding long immersion in sea-water, and are thus 

 suited for oceanic transport. That the present general 

 diffusion of a large proportion of the plants inhabiting 

 the tidal forests and sandy and muddy sea-shores of the 

 tropics is in a great measure due to oceanic currents is, 

 in his opinion, quite certain from the evidence ; a view 

 from which few, we imagine, will be inclined to dissent. 

 In illustration of this subject he gives (taking a small 

 selection of flowering-plants whose seeds are transported 

 by oceanic currents and by birds) the following picture 

 of the gradual invasion of an island by herbs, shrubs, and 

 trees. " The seeds of many almost ubiquitous sand- 

 binding grasses may be reckoned among those which are 

 cast ashore in a vital condition, and we assume that these 

 grasses are amongst the first flowering-plants to obtain a 

 footing. Other herbaceous plants met with in the earliest 

 stage of such an insular flora are Portulaca, Sesuviiiiii, 



Canavalia obtusifolia, and Ipomtea biloba [I. pes-caprce) ; 

 all of which seem to possess an unlimited power of 

 colonisation. Moreover, they provide the conditions 

 necessary for other plants to be able to establish them- 

 selves. Among the early shrubby occupants, Snriana 

 maritima, Pemphis acidula, Scavola Kcenigii, and Tourtie- 

 fortia argeiitea are prominent, being found on the most 

 remote islets of the Pacific and Indian Oceans within the 

 tropical and sub-tropical zones. Where there are muddy 

 shores, there the various mangroves {Rhizophora, 

 Britgiiiera, Aviceniiia, Vitex, &c.) take possession. 

 Among the first real trees are Heritiera littoralis. Hibis- 

 cus iiliaceus, and Barringtonia speciosa, together with 

 screw-pines. After this nucleus of a flora has been 

 formed, it is comparatively easy for other arrivals to 

 establish themselves ; and every addition in a measure 

 helps to provide the conditions for a still more varied 

 vegetation." And he concludes : — " It may be safely 

 assumed, therefore, that if oceanic currents and birds 

 have not been the means of dispersing a large number of 

 species of plants, and it is not certain that they have not, 

 they are certainly the most important agents in stocking 

 islands, for without their action the numerous remote 

 coral islands, at least, would still be utterly devoid of 

 phanerogamic vegetation, and consequently uninhabit- 

 able." 



As a concrete illustration of the influence of these 

 agencies in stocking islands, an analysis of the in- 

 digenous vegetation of the Bermudas is given, which 

 shows that 45 species are chiefly littoral plants, 

 the seeds having been probably conveyed to the 

 island by oceanic currents ; 38 are marsh plants, with 

 small seeds, possibly conveyed to the island in mud 

 adhering to birds, though many may have reached in 

 vegetable drift ; 13 are plants with more or less fleshy 

 fruits, and probably were carried by frugivorous birds, 

 leaving a very small number of species introduced, 

 probably indirectly, by man. 



In a chapter upon the Antarctic flora, and the origin of 

 the vegetation of the islands of the South Indian Ocean, Mr. 

 Hemsley subscribes to the view advanced by Sir Joseph 

 Hooker, and maintained by others, that a former greater 

 land connection in the southern hemisphere is necessary 

 to account for the present distribution of the vegetation 

 in this region, and that a northward migration of southern 

 forms has taken place in the past, and has perhaps hardly 

 ceased. He thinks a greater land connection than 

 Wallace allows must have existed, though the continental 

 extension demanded by Hutton is imnecessary. At the 

 same time he admits that a land connection so great even 

 as Wallace assumes, along with alternations of climate, 

 removes many of the difficulties in the way of accounting 

 for present distribution, and if Thiselton Dyer's hypo- 

 thesis be accepted, that the northern hemisphere is the 

 primary home whence a southward migration of the 

 forms of vegetable life has taken place, he considers a 

 sufficient explanation is obtained. But he dissents en- 

 tirely from Thiselton Dyer's view, holding that " until 

 more conclusive testimony is forthcoming of the former 

 e\\siQv\ct oiProteacex,Eticafypti, &c.,in Europe, we cannot 

 avoid the conviction that they originated in the south." 



Mr. Hemsley points out that the absence of general 

 structural peculiarities in insular plants, and the occur- 



