APPENDIX. 39 



small superior radicle. It differs, however, in the form of the 

 anthers and the pecuHar growth of the calyx in fruit. We must 

 not forget, too, the analogy existing between Disoon and the 

 Selaffinacece, in their hippocrepiform 1-locular anthers, their 

 didynamous stamens, and the structure of the ovary, fruit and 

 seed. These genera, for the reasons above given, are probably 

 more allied to the Scrophulariace(e than to the Vei^benacece, 

 Mr. Bentham's view is probably well-founded, that the true 

 MyoporacecB do not differ ordinally from the Verbenace^Py which 

 is confirmed by the occasional presence of albumen, as I have 

 observed, in the seeds of the latter family. 



Another novel point of structure in the DuboisiecB is worthy 

 of our consideration. The placenta; are adnate to the base of the 

 simple dissepiment, the upper portion of which is membranaceous 

 and marked by a thickened nervure, where, as the ovary enlarges, 

 this part becomes split and separated into two lateral por- 

 tions, attached respectively to the opposite walls of the pericarp : 

 consequently both the berry of Duboisia and the capsules of 

 AnthocerciSj Cyphanthera and Anthotroche are incompletely bilo- 

 cular in the summit, and the dissepiment becomes more or less 

 lunulate, as occurs in several genera of the Goodeniacea. Another 

 fact, at the same time, should be remembered, the great approach 

 in the character of the sestivation of the corolla among the Du- 

 boisie(Ey to that existing in the Goodeniacece. 



There is little to add to the observations already made upon 

 the structure of this genus, except to indicate an error in its ge- 

 neric character, as given in the ' Prodromus ' (DC. x. 191), where 

 the radicle is said to point to the basilar hilum : this is an over- 

 sight ; the radicle certainly points to the base of the seed ; but the 

 hilum, as in Anthocercis and its congeners, is seen upon the ven- 

 tral facCj a little below the middle, in the sinus of its slight cur- 

 vature. Mr. Bentham there points to an error in Endlicher^s 

 interpretation of Bauer's analysis above referred to, wherein the 

 seed is mistaken for the placenta, and a tubercle of the seed for 

 the seed itself : whatever may have been Bauer's intention, I can 

 confirm the truth of the above remarks from my own observa- 

 tion, with this difference, that the areolae represented in the plate, 

 and said to be tubercular rugosities, are in fact deep holl 

 The mode of its inflorescence is very pecuhar, and similar to that 

 of Cyphanthera, 



The following designation of its generic features will be found 

 to be more in accordance with the facts just enumerated : 



Duboisia, R. Br. Prodr. 448; Endl. Gen. n. 3906; DC. Prodr. 

 X. 191. — Char, emendat. — Calyx parvus, campanulatus, sequa*- 

 liter ct breviter 5-dentatuSj mox corollse incremento ssepe 



