354 Transactions of the South African Philoso2Jhical Society. 



Africa as a centre to which America and South Africa stand in 

 relations of almost equal degree, as the following table of distribution 

 will show. The tribe has 6 genera in America, chiefly Tropical, 

 1 in the East Indies, 1 in West Tropical Africa, 1 in Tropical 

 and sub-Tropical Asia, Malay Archipelago, and Australia ; 2 are 

 restricted to Madagascar, 1 to Madagascar and North America, and 



1 to the East Indies and Tropical Africa, and the following South 

 African genera : — 



Melasma, Berg. About 24 species ; distributed in the hotter 

 parts of the world (8 in South x\frica). 



Gerardina, Engl. Monotypic ; Transvaal and Tropical Africa. 



Striga, Lour. About 30 species ; distributed over the hotter parts 

 of Africa, Asia, and Australia (6 species in South Africa). 



Buttonia, McKen. 2 in Tropical Africa, 1 in Natal. 



Sopuhia, Hamilt. 20 species ; distributed over Africa, Madagascar, 

 and India (5 in South Africa, chiefly Eastern and Kalahari Eegion). 



Bopicsia, Presl. 2 species in South Africa, 1 in Socotra. 



Buchnera, L. About 75 species ; distributed over the hotter parts 

 of the world (4 in South Africa). 



Cycnium, E. Mey. About 25 species ; mostly in Tropical Africa 

 (3 in South Africa). 



Bamphicarija, Benth. 10 to 12 species ; 3 in South Africa, the 

 others in Tropical Africa, East Indies, and Australia. 



Harveya, Hook. About 27 species ; 21 in South Africa, the others 

 in Tropical Africa and the Mascarene Islands. 



Hyobanche, Linn. 5 species in South Africa only. 



I may add that Charadrophila, Marloth, and Derniatobotrys, Bolus, 



2 endemic genera, which doubtfully belong to Scrophularicacce, seem 

 to point also to American affinities. 



Thus we arrive at the conclusion, fromphytogeographical considera- 

 tions, that there was a double connection between the South African 

 Flora and the Flora of South America — one by way of Australia, 

 accounting for relations of the Flora of the Cape Province, chiefly 

 with southern South America, and one by way of Tropical xlfrica 

 accounting for the relations of the Palaeotropic South African plants 

 with the Flora of Tropical America. The land-connection which the 

 latter relations imply has been conceded as a reasonable hypothesis 

 by numerous zoologists and geologists. Von Ihering (1893, p. 6), 

 starting chiefly from a consideration of the distribution of fossil and 

 recent fresh-water mussels, points out that with reference to its 

 fresh-water fauna South America can be divided into three regions. 

 The northern is connected with the palaearctic region. Central 

 South America has no trace of relationship with North America, 



