ME. BENTHAM ON LOGANIACEJE. 53 



those artificial assemblages, which, in the present state of our 

 knowledge of plants, we are obliged to interpose between some of 

 the great families, to receive anomalous genera rejected from them. 

 Our natural orders, with all the improvements they have received 

 from the most philosophical of modern botanists, are yet as dis- 

 similar in definiteness of circumscription and apparent conformity 

 to nature, as they are in extent. Some indeed, including the two 

 most numerous of all, are so well characterized as to admit of no 

 doubt. The Cruciferce, Leguminosce, Umbelliferce, Composite, La- 

 biatce, JPalmce, Orchidece, Cyperacece, Graminece, and several others, 

 comprehending two-thirds of the known species of plants, are ad- 

 mitted by all botanists without any variation, and although, amidst 

 the thousands of species comprised in each, there may be some one 

 or two which may offer an exceptional character or anomalous 

 structure, indicating some slight approach to other groups, yet we 

 cannot have the least hesitation as to where to draw the line of 

 demarcation. The Himalayan Megacarpceas, although polyandrous, 

 are still decidedly Cruciferous, not Capparideous. The distinction 

 between Leguminosce and Mosacece, although so difficult to be ex- 

 pressed in words, is yet so clearly defined, that we find no single 

 genus or species ever considered as intermediate, and although the 

 passage from the former into Terebinihacece through Copaifera and 

 Connarus be really more gradual, yet it is still between those two 

 genera that the limits are placed by universal consent ; so are they 

 as irrevocably fixed between the closely allied genera Teucriwn and 

 Vitex, which form the connecting link between Labiatce and Verbe- 

 nacece. The vast orders of Umbelliferce and Composited are equally 

 isolated, notwithstanding the anomalous inflorescences of Hors- 

 fieldia and some others in the former and Xanthium in the latter, 

 which at first sight disguise their characters. The few species of 

 Apostasiece are but anomalous Orchidece, rather explaining their 

 structure than connecting them with any particular order. Cype- 

 racece and Graminece retain their typical structure through all the 

 singular modifications hitherto observed. 



There are other orders again, even amongst the most riumerous 

 in species after the Composites and Leguminosce, which are admitted 

 on all sides to be natural, but upon whose precise limits few 

 botanists can be made to agree, an almost continuous chain of 

 intermediate groups connecting them with adjoining ones. Here 

 the severance has generally been made wherever the links have 

 appeared the weakest; but as these weak points have been variously 

 appreciated by different minds, and no definite standard has been 



