CH. II] INTO NEW AREAS 21 



the purpose, and depend on a small transport due to wind or 

 animals, often only of a few inches. Ritigala, which Avas probably 

 covered with a "dry-zone" flora, but which has apparently 

 existed in its present place since the Tertiary period, has only 

 received 103 "wet-zone" plants in all that time, though the 

 conditions are favourable to them, while Krakatau, with virgin 

 soil, has received 137 in less than thirty years. All the work, 

 whether upon dispersal or upon plant-associations, that has been 

 quoted, goes to show the enormous influence of barriers ; but as 

 the floras of most countries, even of most islands, do not show 

 any such influences of the barriers that cut them off, the natural 

 inference is that in general they received the bulk of their floras 

 when the barriers were not there. 



Looking at the dispersal mechanisms in a general way, one 

 gathers a broad impression that they are really of much less 

 importance to plants than one has been inclined to imagine. This 

 is confirmed by the fact that one finds many genera with little 

 or no mechanism for dispersal just as widely spread and cosmo- 

 politan as others with the most perfect arrangements. For 

 example, among the former we find Callitriche, Ceratophyllum, 

 Carex, Cocculus, Desmodium, Evphorbia, Hij^puris, Jwicus, 

 Lemna, Piper, Pistia, Polygonum, Salvia, Utricularia, etc. Al- 

 together more than half the cosmopolitan genera have no good 

 dispersal mechanism. (Cf. Lantana and Tithouia mentioned 

 above, p. 17.) 



Of genera occurring in both Old and New Worlds, the family 

 Avith most (97) is Gramineae, whose fruits are to some degree 

 suited to wind dispersal, but it is followed by Leguminosae (79) 

 which are ill-suited to rapid spread, except to some extent by 

 currents. These families are followed by Compositae, Orchida- 

 ceae, Rosaceae, Rubiaceae, Scrophulariaceae, Liliaceae, Umbel- 

 liferae, Cyperaceae, Cruciferae, Caryophyllaceae, Ericaceae, 

 Euphorbiaceae, Ranunculaceae, Acanthaceae, Convolvulaceae, 

 Coniferae, Labiatae, and Malvaceae, in the order named. The 

 general impression is not that of the predominance of plants 

 with good dispersal mechanisms. 



The first ten largest families in the world (judged by number 

 of genera) — the Compositae, Orchidaceae, Leguminosae, Rubi- 

 aceae, Gramineae, Asclepiadaceae, Euphorbiaceae, Umbelliferac, 

 Cruciferae, and Acanthaceae — are not remarkable for the pos- 

 session of extra good methods of dispersal, excepting the first 

 two. Yet not only have they the largest number of genera in 



