MB. G. BENTHAM OS EUPHORBIACEJK. 247 



ever a Baloghia, an Adenopeltis, or a Homalanthus in Bohemia, con- 

 trary to all presumptive evidncee. Not a single fruit elastically 

 tncoccous has to my knowledge been found amongst these remains; 

 and if there were, it would be hard to distinguish their fossil 

 impressions from those of some Eutacese or Simarubea?. Origin, 

 therefore, can only be deduced conjecturally from peculiarities in 

 the modern geographical distribution. 



The first and most striking feature is evidently the tropical nature 

 ol the Order. The few extratropical tribes or genera (except pos- 

 sibly Buxese) would appear, as in the case of Mimoseae, to have been 

 independently derived from tropical ones, although further diffe- 

 rentiated and multiplied in temperate regions. "We observe no evi- 

 dences of ancient communication between distant extratropical 

 regions (such as that between South Africa and Australia, exempli- 

 fied inRestiacese, ~Proteace&, Helichri/sum, Diosmeae, etc.), no west- 

 ern Old-World connexion of the temperate floras (as exhibited 

 between western South Africa and western Europe in the case of 

 vxesniteae, Erica, Lobelia, etc.), no West-American connexion be- 

 tween California and Chile, no group common to Mexico, extratro- 

 pical South America, and South Africa — the only approach to such 

 a connexion being that of JDysopsis with the Stenolobese, to which 

 1 shall presently recur ; and the northern extratropical connexion 

 is very rare. On the other hand, the connecting links between the 

 extratropical tribes and the tropical ones will generally be found 

 within the tropics, as I shall endeavour to show when considering 

 the geography of special genera. 



If there are strong arguments in favour of the tropical origin 

 of the order, the question of what region within the tropics gave 

 hirth to it is not so easily solved. Excluding the southern extra- 

 tropical Stenolobese, all the five tribes are common to the New 

 and the Old World ; about 110 genera, or marked sections, are 

 exclusively Old World and mostly tropical, from various parts of 

 w hat has been supposed to be a once continuous tract of terri- 

 tory from tropical Africa to the Malayan archipelago and South - 

 Pacific islands ; about 60 genera are exclusively American, the 

 tropical ones mostly limited to the eastern region from South 

 Brazil to the West Indies and Central America. About 17 are 

 common to both continents : of these 8 are chiefly American with 

 American relations, 2 are represented in the two continents by 

 distinct but closely allied sections, 1 is an Old- World genus but 

 slightly represented in America, the remaining 5 or 6 may be 



