594 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



areas; and, finally, he seems to have come to the same conclusion 

 as that which he reached in the classification of the vegetable king- 

 dom adopted by him in the monumental work which he produced 

 in collaboration with Bentham, the " Genera Plantarum " (3 vols., 

 octavo, 1862-1883). This conclusion was that, while we are still 

 seeking a closer knowledge of the phyletic connections of the floras 

 and faunas of the world, it is, in view of practical purposes (that is 

 to say, for facilitating the accumulation and orderly arrangement 

 of our knowledge), better to adopt a frankly arbitrary series of 

 groups and provinces agreed upon and accepted because they are 

 traditional and serviceable for purposes of reference, than to as- 

 sume prematurely that we are in a position to define the limits and 

 connections of all natural phyto-geographical provinces and of all 

 phyletic groups. To do this we have not yet (he thought) sufficient 

 knowledge, though we already see clearly much of the outlines and 

 the needful lines of inquiry. 



The means and the causes of the migration of plants were matters 

 of extreme importance in the great problem of distribution and the 

 closety connected problem of the changes of land and water on the 

 earth's surface. These were the subject of speculation and inquiry 

 by both Darwin and Hooker. Hooker had at first put forward the 

 hypothesis of a lost circumpolar continent in order to account for 

 the facts of plant distribution in the southern hemisphere. But 

 Darwin favored the view of the persistence even from Silurian 

 times of the great continental masses at present existing, and the 

 radiation from the northern temperate and subarctic region of 

 successive floras by spreading along the cold mountain chains which 

 extend through the tongue-like southward projections of continental 

 land — to-day traceable as South America, Africa, and Indo-Malaya. 

 Transport of seeds, etc., by ocean currents, by wind, and by birds 

 and other such agencies was shown experimentally by Darwin to be 

 possible in many cases, but the emergence and submergence of large 

 tracts of land as bridges or connections across the deep ocean beds 

 were rejected by him. Hooker writes to Darwin in 1881 : 



Were you not the first to insist on this [the permanence since the Silurian 

 period of the present continents and oceans], or at least to point this out? Do 

 you not think that Wallace's summing up of the proof of it is good? I know 

 I once disputed the doctrine or rather could not take it in ; but let that pass. 

 (Vol. ii, p. 224.) 



He goes on to say, in reference to the address which he was pre- 

 paring for the British association meeting at York, in which after 

 many years' labor he expressed his final conclusions on geographical 

 distribution : 



I must wind up with the doctrine of general distribution being primarily from 

 north to south with no similar general flow from south to north — thus support- 



