Mr. J. Miei's on several genera hitherto jjlacedin Solanacese. 171 



enecl placentiferous dissepiment, its many-seeded capsular fruit, 

 and its slightly curved embryo with an inferior radicle, are cha- 

 racters quite opposed to its admission into that family. Nor can 

 these be made to harmonize either with the Scrophulariacea or 

 Solanacece, to the latter of which they offer a nearer affinity. 

 These characters are sufficiently prominent and distinct, and de- 

 mand a more attentive investigation. 



7. Anthotroche. — This genus was placed by Prof. Endlicher 

 in Sa^ophulariacecB, among the Salpiglossidece , but it has been 

 since excluded from the order by Mr. Bentham, and referred to 

 Solanacece (DeCand. Prodr. x. p. 586). It appears to me how- 

 ever to have as little relation with the one as with the other of these 

 families. Upon examining a specimen belonging to this genus 

 from Swan River, I find that in the structure of its anthers it 

 agrees entirely with that just described as existing in Duboisia ; 

 this consists of one reniform unilocular cell, fixed extrorsely on 

 the filament, and dehiscing on the exterior face by one hippocre- 

 pical suture. Here the tube of the corolla is short and straight, 

 and the border is divided into five regular lobes, which are ro- 

 tately expanded ; the stamens are 5 and equal. The ovarium 

 has an epigynous prominent stylobasic gland as in Cacabus, ana- 

 logous to that of Hyosci/amus. 



Respecting the Duboisiece it only remains to be observed, that 

 the main points of distinction between it and the other tribes 

 with which it is here associated, will be found to exist in the ex- 

 trorse direction of the anthers and the singular aestivation of the 

 corolla, peculiarities which, although very remarkable, are not 

 of themselves of sufficient importance to claim for the plants that 

 compose it the rank of a separate family, but they constitute a 

 very distinct tribe of the AtropacecB. It will consist of two sec- 

 tions : 1. Euduboisiea, with baccate fruit, and 2. Anthotrochea, 

 with capsular fruit, comprising Anthocercis and Anthotroche. It 

 corresponds with the other tribes of the Atropacece in the ori- 

 gin of the floral peduncles being lateral with respect to the point 

 of insertion of the petiole. 



4. Schizanthece. — The genus Schizanthus, from the lateral 

 extra-axillary insertion of its pedicels and other characters, ap- 

 pears evidently to belong to the Atropacece rather than to the 

 Scruphulariacece, but it does not accord with any of the tribes 

 above noticed. It differs from them in the structure of its an- 

 thers, which consist of two parallel cells, quite distinct and sepa- 

 rated from one another, but conjoined by a broad membranaceous 

 connective, upon which they are dorsally attached : it possesses 

 five stamens, of which three are quite anantherous and rudimental; 

 the corolla is deeply cleft into niimerous unequal segments which 

 have an imbricate sestivation. Its stigma approaches the form 



