252 MR. G. BEtfTHAM OX ETJPHOKBIACEJE. . 



where various races are descended from a single one, all may in 

 their further differentiation show variations in some characters, 

 whilst one only may assume some special character, and that this 

 may be continued through a number of secondary races, which 

 may in other respects vary in the same manner as the descendants 

 of such primary races as had not assumed the special character. 



The fact above referred to, that the nearest approach to the 

 special character of StenolobeaB has been observed in the extra- 

 tropical South-African genera Adenocline and Seidelia, is not 

 sufficiently definite to afford any speculations as to derivation 

 unsupported by other data, no close connexion between extra- 

 tropical South -African and Australian Euphorbiacese being on 



record. 



3. Bttxilx. 



Here, as in Stenolobeae, the constancy of a specially excep- 

 tional but^essential character seems to indicate a community of 

 origin in plants otherwise very different in structure ; but that 

 community must have been exceedingly remote. The genera are 



wi 



exceptions in Buxus itself, consisting of solitary or few very 

 distinct species of very limited areas — all evidences of very ancient 

 but expiring races. All have a northern character, no trace of them 

 appearing south of the tropics. Buxus, the only genus with 

 widely spread variable species, and thus still in a flourishing state, 

 may, however, be as ancient as any of them, and possibly the 

 nearest to the common parent of the tribe. It must have existed 

 very nearly in its present form at that very remote period when 

 the warmer regions of the Old and the New World were in con- 

 nexion or communication. In the former it has left distinct spe- 

 cies of limited areas in east tropical Africa and Madagascar, and 

 perhaps thence or from further north may have spread in equally 

 remote times over the northern hemisphere in widely extended and 

 variable forms. In America it has assumed a slightly (but very 

 slightly) different form, and is now limited, in a small number of 

 species, to the West Indies, without any evidence of its having 

 ever extended into North America, or having travelled by that 

 route from the Old World. Sarcocoeca, in some respects very near 

 Buxus, yet separated from it by a well-marked gap, is still flourish- 

 ing in the mountains of tropical Asia, and may have formerly 

 extended thence northward, becoming modified into the Japanese 



