Mr. J. Miers on the Affinities of the Olacacese. 131 



I shall also be able to add some new evidence confirmatory 

 of the observations of Dr. Arnott relative to the structure of 

 Gruhbia (Hook. Journ. Bot. iii. 266) ; its alliance however ap- 

 pears to me nearer to the Hamamelidacece than the Bruniacem ; 

 and the tribe of the Ophirice including Grubbia and Ophiria 

 (Strobilocarpus) may well form a sectional division of that order, 

 the limits of which will require some modification to include this 

 and the Diclidantherece as new and distinct tribes. 



The structure observed in Humirium confirms what I urged 

 in regard to the nature of a stipitate torus {lac. cit. p. 176), and 

 of that of the cupuliform disk so frequently alluded to in former 

 pages. We there perceive an ovarium perfectly free and sup- 

 ported on a distinct gynophorus : this is surrounded at its base 

 by a conspicuous cupuliform ring, toothed on its margin, but 

 perfectly free, on both surfaces, down to the base : the ovarium 

 is hairy in all parts, except in the basal portion inclosed within 

 that cup, but not the slightest adhesion exists, either with its 

 glabrous portion or with the gynophorus. Outside of this 

 hypogynous cup is seen another cupshaped ring, serving to 

 support the stamens, which in H, floribundum is entire, smooth, 

 and fleshy outside, and supports the many series of filaments 

 upon its margin as well as upon the whole of its inner face, 

 forming thus an annular ring, free both from the hypogynous 

 cup and the petals. Here therefore we perceive the gynophorus, 

 cupuliform disk, staminiferous cup, petals, and sepals, each a 

 distinct development, and each free to the base, but all springing 

 from a fleshy torus which is simply an expansion of the apex of 

 the pedicel. The torus, therefore, as an organ well marked in 

 many of the Thalamiflorce, must not be confounded with any of 

 the developments which it serves to support*. 



Many of the inferences drawn from the numerous facts in- 

 dicated in " the Remarks on the Affinities of the Olacacese " {loc. 

 cit.) are so much at variance with long-established opinions, that 

 I cannot expect they will at first be favourably entertained. Ex- 

 perience has shown, when conclusions upon erroneous grounds 

 have once been made by high authority, and these confirmed by 

 every subsequent author, that nothing short of actual demonstra- 

 tion, and that of the most positive character, can establish other 

 and more correct inferences. In the " Remarks " alluded to, 

 considerable doubt has been thrown on the deductions of some 

 of the most eminent botanists, — men celebrated for the general 

 accuracy of their observations, and for the soundness of their 

 views regarding botanical affinities : I should therefore incur the 



* See many excellent remarks, all tending to the same conclusions, in 

 two chapters on the Disk and Floral Receptacle, in Aug. St. Hilaire's * Le- 

 90ns de Botanique,' p. 455-466. 



9* 



