Mr. J. Miers on the genus Schwenkia. 187 



a portion of the glandular-looking lobes (true segments) of its 

 border in Schwenkia, must be considered one of those exceptional 

 cases which are occasionally met with in a great many orders ; 

 it serves as a point of osculation between the Solanaceee and 

 Scrap hulariacea, in which latter family, the want of symmetry 

 in its parts, and a total or partial suppression of one of its 

 stamens, form almost universal characters. On the other hand, 

 we meet in the same family several cases where the corolla is 

 pentamerous, and as regularly symmetrical in its parts as in 

 Solanacece ; thus in Capraria {Xuaresia, R. & P.), we find a 

 corolla with a border of five equal lobes and five equal stamens ; 

 so also in some species of Verbascum, and in Sibthorpia, where 

 likewise the stamens are generally five, and equal in number to 

 the regular segments of the border, although rarely four or eight 

 occur. In my definition of Schwenkia, as given below, I have 

 modified somewhat Mr. Bentham's view of the structure of the 

 corolla, considering the expanded segments of the border to be 

 analogous in their nature to the corona of Nectouxia. 



Referring to the question of aestivation, it will be seen that in 

 the sections Chcetochilus, Euschwenkia, and Brachyhelus, where 

 the segments of the corona are small, they are valvately con- 

 joined in bud by their floccose margins into a short cone, that closes 

 the mouth of the tube, the lobes lying over them, and pointed 

 toward the axis : in Brachyhelus, these lobes, which are several 

 times longer thanthetoothed segments, soon become approximated 

 in the axis, where they are connately disposed in an erect central 

 column, so that both lobes and segments may be said to have a 

 valvate sestivation : in Cestranthus the lobes are reduced to short 

 teeth, but the segments of the corona are of considerable length, 

 linear and acute, and also valvately disposed in bud, into a 

 lengthened pentangular cone, exhibiting at its basal angles the 

 five short lobes, as so many salient erect points. In Cardiomeria 

 the lobes are equally short and similarly situated, but the very'broad 

 emarginated segments have their margins valvately disposed, and 

 they are replicated lengthways down the middle, as in Datura, 

 so that the corolla appears in bud like a slender tube swollen 

 into a pentapterous form above, and terminated by five semi- 

 lunate wings, depressed at the point of their union in the axis, 

 and furnished on the external angles with the five salient short 

 erect lobes, like so many uncinate teeth. 



Besides the considerations above described, we have the evi- 

 dence in the structure of the seed, that this genus must be 

 referred to Solanacece, and not to Scirophulariacea, because the 

 embryo, which is slightly curved, has its radicle pointed to the 

 basal angle of the seed, and turned away from the ventral hilum, 

 as shown by Gaertner (De Fruct. tab. 214), while in the latter 



