690 BEPOEX— 1888. 



are scattered over the mountainous regions of Central Africa, and whicli survives 

 i n its greatest concentration at the Cape. 



The American botanists at Harvard are still systematically carrying on the 

 work of Torrey and Gray in the elaboration of the flora of Northern America. The 

 Russians are, on their part, continually adding to our knowledge of the flora 

 of Northern and Central Asia. The whole flora of the North Temperate Zone can 

 only be regarded substantially as one. The identity diminishes southwards and 

 increases in the case of the Arctic and Alpine regions. A collection of plants brought 

 us from high levels in Corea by Mr. James might, as regards a large proportion of 

 the species, have been gathered on one of our own Scotch hills. 



We owe to the munificence of two English men of science the organisation of 

 an extensive examination of the flora and fauna of Central America and the publi- 

 cation of the results. The work, when completed, can hardly be less expensive than 

 that of the results of the ' Challenger ' voyage, which has severely taxed the liber- 

 ality of the English Government. The problems which geographical distribution 

 in this region present will doubtless be found to be of a singularly complicated 

 nature, and it is impossible to overestimate the debt of gratitude which biologists of 

 all countries must owe to Messrs. Godman and Salvin when their arduous under- 

 taking is completed. I am happy to say that the botanical portion, which has been 

 elaborated at Kew, is all but finished. 



In South America, I must content myself with referring to the great ' Flora 

 Brasiliensis,' commenced by Martins half a century ago, and still slowly progressing 

 imder the editorship of Professor Urban at Berlin. Little discussion has yet been 

 attempted of the mass of material which is enshrined in the mighty array of volumes 

 already published. But the travels of I\Ir. Ball in South America have led him to the 

 detection of some very interesting problems. The enormous pluvial denudation of 

 the ancient portions of the continent has led to the gradual blending of the flora of 

 different levels with sufficient slowness to permit of adaptive changes in the process. 

 The tropical flora of Brazil, therefore, presents an admixture of modified temperate 

 types wJiich gives to the whole a peculiar character not met with to the same 

 degree in the tropics of the old world. On the other hand, the comparatively 

 recent elevation of the southern portion of the continent accounts, in Mr. Ball's eyes, 

 for the singular poverty of its flora, which we may regard indeed as still in progress 

 of development. 



The botany of the ' Challenger ' expedition, which was also elaborated at Kew, 

 brought for the first time into one view all the available facts as to the floras of the 

 older oceanic islands. To this was added a discussion of the origin of the more 

 recent floras of the islands of the AVestern Pacific, based upon material carefully 

 collected by Professor Moseley and supplemented by the notes and specimens 

 accumulated with much judgment by Dr. Guppy. For tlae first time we were enabled 

 to get some idea how a tropical island was famished with plants and to discrimi- 

 nate the littoral element due to the action of oceanic currents from the interior forest 

 almost wholly due to frugivorous birds. The recent examination of Christmas 

 Island by the English Admiralty has shown the process of island flora-making in 

 another stage. The plants collected by Mr. Lister prove, as might be expected, to 

 be closely allied to those of Java. But the efi'ect of isolation has begun to tell ; 

 and I learn from my colleague, ProfessorOliver, that the plants from Christmas Island 

 cannot be for the most part exactly matched with their congeners from Java, but 

 yet do not diff"er sufficiently to be specifically distinguished. We have here, there- 

 fore, it appears to me, a manifest case of nascent species. 



The central problem of systematic botany I have not as yet touched upon: this 

 is to perfect a natural classification. Such a classification, to be perfect, must be 

 the ultimate generalisation of every scrap of knowledge which we can bring to bear 

 upon the study of plant affinity. In the higher plants experience has shown tliat 

 we can obtain results which are sufficiently accurate for the present without carrying 

 our structural analysis very far. Yet even here, the correct relations of the 

 Gymnosperms would never have been ascertained without patient and minute 

 microscopic study of the reproductive processes. Upon these, indeed, the correct 

 classification of the '\'ascular Cryptogams wholly depends, and generally, as we 



