ME. G. BENTHAM ON TILIACE^E. 53 



and Elceocarpeee ; the latter considered by some as a distinct order, 

 founded on the notched or divided petals and the dehiscence of 

 the anthers ; whilst TUiece have been subdivided into two tribes, 

 the Sloanidece Avithout petals, and the Grewidece with petals. To 

 me it appears, however, that Sloanea is much nearer allied to 

 Elceocarpus than to Grewia ; and it is probably under the same 

 idea that Grisebach proposes to take the anthers as the primary 

 character for the division of the order. But here, again, Apeiba 

 and Sloanea, which are so different in other respects, are brought 

 next to each other, and removed from other genera with which 

 they appear to be generally much closer connected. If, however, 

 we take the nature of the petals in the first instance, and secondly 

 that of the stamens and calyx* and, following out the principle 

 that governs the demarcation of the order, if we take neither of 

 these characters as absolute, and admit of more or less exceptional 

 genera, we have an apparently more natural distribution of the 

 thirty-seven genera which we propose to adopt, into six tribes, 

 grouped into two suborders. 



The difference between these two suborders, to which we may 

 give the names of Holopetalce and Heteropetala, would then be in 

 some measure analogous to that which we have established between 

 Bixaceee and Samydacece, but much less marked. In the Holope- 

 tales the petals are always present (except in one or two of the 

 numerous species of Greivia) ; and, although sometimes small, they 

 are of a true petaloid nature, entire or very slightly notched, 

 glabrous, narrowed at the base, and deciduous. The stamens are 

 decidedly hypogynous in all but Muntingia, inserted either on a 

 small torus immediately within the petals, or on a raised torus, 

 round the base of which are placed the petals. In the Hetero- 

 petalce the petals, when present, are of a more or less calycme 

 nature, often notched, fringed, or divided, hoary or pubescent 

 outside, attached by a broad base and usually persistent, so as to 

 have been sometimes described as inner calycine segments ; and 

 in some genera they are entirely absent, or present only in a very 

 few species. The disk (usually covered with stamens quite to the 

 ovary) is usually thick— in some genera so broad and flat as to 

 appear perigynous, in others raised, bearing the stamens on the 

 top, and surrounded by the petals at the base, very rarely small 

 and hemispherical. 



These two suborders would again be subdivided into six tribes, 

 more or less natural, of which the KolopetaU would comprise 

 four.— 1. Brownloioiece, marked by a gamosepalous calyx, and 



