218 MR. G. BEKTHAM ON EUPHORBIACE-E. 



considered as pretty generally belonging to both. When it is 

 further considered that the great majority of isolated or remotely 

 connected monotypie or small genera, apparently remains of 

 ancient expiring types, belong to the Old World, although a few 

 such may be also f onnd in America, and that the tendency to a 

 multiplication of species is greater in America than in the Old 

 World, we may be led to conjecture that the most ancient home 

 of the Order was in the Old World, but that several of the princi- 

 pal forms were differentiated and widely spread before that remote 

 period when the present impassable barriers opposed by the At- 

 lantic and Pacific did not exist, or were crossed over in some 

 manner of which no plausible explanation has been suggested. 

 It would also appear that interchange of forms between the prin- 

 cipal Old- World centres of differentiation, tropical Africa and 

 Madagascar on the one hand, the Malayan and South-Pacific 

 islands on the other, continued long after the interposition of the 

 obstacles preventing the spread of the new T American forms. In 

 elucidation of these conjectures it may be well to go through 

 seriatim the different tribes and principal genera with reference 

 to their actual distribution, premising, however, that, owing to 

 our imperfect acquaintance with a large number of tropical fru- 

 tescent or arborescent genera, the best tests of genetic history, 

 much may have to be modified hereafter, both in the data given 

 and in the conclusions drawn from them. 



1. Euphorbie^. 

 The two small genera Anthostema and Synadenium probably 

 represent a very early stage of the tribe — that is, a separation 

 from the common stock before the involucre had become so de- 

 finitively consolidated and the perianth so completely annihilated 

 as is uniformly the case in Euphorbia. They are both African, 

 one only extending into Madagascar ; and both are limited to two 

 or three species, neither of them very common, nor showing 



fprogressiv 



much variation — indications rather of expiring than < 

 races. Calycopeplus, another genus of tw r o or three species only 

 is an Australian offset, allied to the above two ancient African 

 genera, and probably differentiated at a similar remote period : 

 the structure of the flower-head is much like that of Synadenium, 

 except that the perianths are perceptibly developed, and the in- 

 volucral glands less so ; and in adaptive characters it has assumed 

 a habit not uncommon in Australia. This genus and the Euphor- 



