CHAPTER XXIII. 



THE BOTANICAL REGIONS OF SOUTH AFRICA. 



Reference has been made to several orders \vhich are found only in 

 South Africa and which are confined wholly or largely to the south- 

 western portion. They probably represent very ancient types of plants 

 of southern origin which have been confined in this restricted area by 

 impassable climatic and geographical barriers farther north. 



The origin of South African flora has greatly interested botanists 

 since Sir Joseph Hooker published the results of his observations during 

 the Antarctic expedition with Sir James Ross. 



Certain orders of South Africa are abundantly represented in 

 Australia ; owing to the predominance of species in the latter country, it 

 has apparently been the centre of distribution for several orders found in 

 both South Africa and South America. 



A South African tree Mctrosidcros belongs to the order Myrtaceje 

 which supplies us with Australian blue gums. This tree, found in the 

 western mountains of Cape Province, has its centre ol distribution in New 

 Zealand. Besides the South African species, M. avgustifolia, Sm., one 

 species is found each in Chili, Java, and Madagascar. 



ProteaceEe are especially interesting as representing a southern dis- 

 tribution. Of the two main groups into which the order is divided (see 

 p. 285) the first is found in Australia and South Africa, while the 

 GreviUoidea; are largely represented in Australia and Chili, but none 

 occurs native to South Africa. Many of the South American types thus 

 associated are in the extreme south-western part of Chili. 



A glance at the map will show that South America is much nearer to 

 Africa than it is to Australia, it will therefore appear strange that its 

 flora should be more closely related to Australia. 



Among other orders representing a South African-Temperate 

 Australian alliance may be mentioned Irtdaccif, Rcstiaceir, HatiiaJor- 

 aceiT^ Composites, ThymeliacCiF, and Rubiacea-. 



By referring to the chart, p. 375, a partial similarity will be seen 

 between the distribution of Proteaceas and Cycadaceae. Fossil forms 

 of the latter order show that they were once widely distributed in the 

 northern hemisphere. Well-grounded evidence for fossil Proteacea; in 

 the northern regions is wanting. 



A distribution such as that of Proteace;e, supports the belief that 

 a former land connexion existed between South .\frica and Australia, 



374 



