OF ARTS AND SCIENCES. 37 



paratively few, and very thick-walled pyreno?, amply distinguisli Bohea 

 from Timonms. From the osstivation, the two genera would fall into 

 different subtribes, if we implicitly follow the distribution of the Goffe(S<s 

 suggested by Mr. Bentham, who, in the Niger Flora, and in the Kew 

 Journal of Botany, has brought out, with his usual sagacity, the best 

 characters for a natural arrangement of the RuliacecB. Timonius, with 

 valvate jestivation, would fall into his Vangueriece, and Bohea into his 

 Guettardece. But characters from the aestivation of the corolla, conven- 

 ient and generally reliable as they are, must in this, as in many other 

 cases, give way to other considerations. And the close similarity of these 

 two genera in most other respects, especially in their nearly exalbumi- 

 nous embryo, the plug-shaped funiculus filling the upper part of the 

 cell, as also the delicate venular meshes of the leaves (of which traces 

 occur in most GhomelicB and G-ueUardce), plainly requires us to refer 

 Timonms also to the Guettardece. This is well confirmed by Wight 

 and Arnott's genus Eupyrena, which is exactly intermediate between 

 • Timonius and Bohea, having the corolla, &c. of the former, and the 

 ovary and fruit of the latter.* A leading character of the Guettardece, 

 as I should define the group, has escaped general notice, namely, the 

 wholly or nearly exalbuminous embryo, the embryo being a macro- 

 podous radicle with small and obscurely-marked cotyledons occupying 

 merely the lower extremity. This, as well as the true pericarpic di- 

 rection of the embryo, however, was rightly understood by Grertner, in 

 Guettarda speciosa. What A. Richard calls a thin fleshy albumen in 

 this plant, is apparently the tegmen of the seed ; and what he took 

 for the radicular extremity of the embryo (" cet embryon est dresse ") 

 is the cotyledonar end. The mistake is the more remarkable on 

 Richard's part, since he had recognized the crustaceous plug at the 

 summit of the cell as the funiculus. Otherwise, the tapering of the 

 ovule and young seed to an acute apex at the base of the cell might 

 readily lead one to regard them as attached there, and may have given 

 origin to the character " semina erecta," assigned by De CandoUe and 

 others, and Avhich even Miquel leaves in the generic character (with a 

 mark of doubt), although, following Bentham, he rightly defines the 



* The fruits of Eupyrena, with which I have been supplied by Dr. Hooker and 

 Dr. Harvey, do not furnish sound seeds ; but I have no doubt that these are exal- 

 buminous or nearly so. The ovary is sometimes only quinquelocular ; and the 

 ovules are suspended. "Semina erecta" is a phrase wi'ongly introduced into the 

 generic character by Endlicher. 



