PUESIDENTIAL ADDRESS — SECTION C. 8l 



science ; for this purpose a handbool>: carefully and accurately 

 prepared, if it covers most of the species of a flora, will be of 

 great assistance, even though many of the rarer forms are 

 omitted. To accomplish the best results in the shortest time, 

 therefore, I would suggest the adoption of two courses : The 

 collection of (i) the dominant, sub-dominant, and social mem- 

 bers of the vegetation in the allotted territory; (2) fairly com- 

 plete collections from selected localities which from their , 

 topography, soil or climate suggest a rich and varied 

 flora. With regard to the first of these suggestions, I 

 have found that a reconnaissance, or sort of flying 

 survey, of the territor}^ under review enables one to divide the 

 region roughly into two or more distinct phyto-geographical 

 areas. 1 have adopted the plan of taking one locality (for con- 

 venience a single registered farm of, say, approximately ten 

 square miles in extent ) and collecting, as far as possible, every- 

 thing found growing upon it, including the cultivated i)lants. 

 This gives surprisingly interesting results. 



Geographical Distribution. 



A branch of botanical survey work which has proved most 

 fascinating has been the geographical distribution of species of 

 plants. This is of value in connection with our study of plant 

 ecology and phylogeny. 



Although a good deal is now known of the floras of the 

 Eastern and Western Regions, Natal, the Orange Free State 

 and the Transvaal, there are large areas as yet almost unvisited. 

 Thanks to the energy of the late Professor Pearson, a former 

 President of this Section, we are now in possession of valuable 

 data from the hitherto but little known though very interesting 

 region of Namaqualand. ■ But we know surprisinglx- little of 

 the flora of the great Drakensberg Range ; Basutoland is almost 

 unknown ; the same may be said of that huge tract of country 

 from the Komatie River to the Limpo|x>, lying between the 

 Drakensberg and the Lebomlx) Mountains, and particularly that 

 great unsurveyed tract of Government land occupying the ex- 

 treme north-east corner of the Transvaal ; and the whole of 

 the Limpopo drainage basin falls into the same category. Very 

 little is known, either, of .Zululand, the Transkei, and the western 

 portion of the Kalahari. It may well be said that there is no 

 ])ortion of South Africa — unless perhaps it is the Cape Penin- 

 sula — of the flora of which we can claim a fairly accurate 

 knowledge. Even such old collecting centres as Uitenhage are 

 yielding quite distinct ne\\ species, while Conrath's small col- 

 lection at jModderfontein. near Johannesburg, has yielded such 

 a number of unnamed though well-known forms as to suggest 

 that we may have long been in error in ascribing to Transvaal 

 species, names from the " Flora Capensis " which really 

 belong to species restricted in distribution to the Cape Province, 

 or in a few cases even to Europe. 



D 



