LINXEAX SOCIETY OF LONDON. 67 



the large genera have a small projwrtioii of eudemics, e. g., Cro- 

 talana, with thirty-three species, and Tephrosia, witli thirty-one, 

 have only eight and nine endemics respectively. 



The species show great variation in range, even in the same 

 genus; some range almost the length of the Continent ; others are 

 restricted to very limited areas ; every possible variation of ranc^e 

 between these two extremes is covered by the maiority of the 

 species. 



Classified according to their geographical range of distribution, 

 the 1 apdionacea) tall into five very distinct groups : (1) the South- 

 western Cape Province Element, with only five species. AVhen 

 we take into further consideration the fact that eleven endemic 

 bouth-western Oupe genera of Papilionacete, with 270 species, do 

 not occur at all in the Transvaal, it is clear tliat the connection 

 between the floras of the Transvaal and the South-uest Cape 

 Province is negligible. 



(2) The Kalahari Element. This comprises only nine species 

 conhrming a much earlier- formed conclusion that the inclusion of 

 the Iransvaal m the - Kalahari Region" of the later vohnnes of 

 tlie IMora Capensis ' is quite misleading. 



(;}) The Rain-forest Element of the eastern higl, mountains 

 witli only about five species. 



These three elements together comprise not more than 6 per 

 cent, ot the total Papilionaceous flora. The remaining 94 per 

 cent. (oO(! species) are divided between (4) the Tropical African 

 Ele.nent with 1G7 species (51 per cent, of the total), and (5) the 

 Warm lemperale Plateau or the high-veld grass stepi^e flora of 

 the eastern Transvaal and Orange Eree State, and uplands of 

 Natal, Gr.qualand East, Pasutoland and the eastern portion 

 of the Cape Province. This Warm Temperate Plateau Element 

 comprises i:^9 species or 43 per cent, of the total. 



Of these ;^0(i species comprising the two predominant elements, 

 123 are endemic to the Transvaal. These endemics form 38 per 

 cent, o tlie total; this proportion appears low when compared 

 with the /2 per cent, endemics in New Zealand, or the 82 per 

 cent ot the Hawah-an Islands. But if we add to the species which 

 are strictly endemic within the political boundaries of the Trans- 

 vaal, tho.se species which range into the Border States but not 

 beyond them, the number of endemics in the thus enlarged area 

 IS /8 per cent, of the population. In other words, if the Border 

 S ates were to be submerged to-day, leaving the Transvaal as an 

 island we should have 78 per cent, of its Papilionaceie as endemics, 

 and 22 per cent, would be " wides " occurring also on the mainland. 

 Ihus the only reason that we have such a small percentaoe of 

 endemics ,n the Transvaal to-day is that some of them range 

 across the political boundaries into Border States, while in the 

 case ot islands, the oceans have formed an insuperabh. barrier 

 either preventing the spivad of species evolved since the isolation 

 occurred, or destroying the individuals which had spread before 

 tlie submergence ot outlying territory. 



/2 • 



