107 



the real stamens ; because in Reseda alba, and some others, in which a 

 union of filaments takes place in the perfect floret, there is a corresponding 

 but more cumplete union of the sterile appendages; and because occasion- 

 ally in Reseda odorata, stamens are changed into bodies altogether similar 

 to tlie sterile appendages ; and in Reseda Phyteuma the same appearance 

 is always assumed by the perfect stamens after the anthers have performed 

 their functions. Thirdly, That there is an equal analogy between the calyx 

 of the neutral florets and that of the perfect floret ; because both have a 

 peculiar glandular margin, the same form, both produce their stamens from 

 their surface ; and because the upper edge of the calyx in the sterile florets ' 

 has the same relation to the axis of each particular head as that of the 

 perfect floret has to the axis of the whole inflorescence. In Reseda Phy- 

 teuma, which has the margin of its neutral florets rolled back, the same 

 thing occurs in the perfect floret. Fourthly, That there is no instance of 

 the same analogy existing between the disk and petals of other plants." 

 Coll. Bot. no. 22. Hence I inferred that the genus must be excluded from 

 even the vicinity of Capparidese, with which it is usually placed. This view 

 of the structure of Reseda, however paradoxical it may appear, has been 

 adopted by M. Decandolle; but Mr. Brown, in the Appendix to Major 

 Denharns Narrative, has advanced various arguments in opposition to it. 

 By these I was at first induced to believe that I was mistaken in my theory ; 

 but upon reflection, and a subsequent repetition of the observations I origin- 

 ally made, 1 have been led to decide that Mr. Brown's arguments, strong as 

 they undoubtedly are, do not carry conviction with them, and are, in fact, less 

 weighty than they seem to be. In the first place, this learned botanist does 

 not attempt to invalidate some of the arguments upon which I was led to my 

 original conclusion ; and secondly, those which he has advanced in support of 

 the contrary opinion appear to me to be open to objection. Mr. Brown's 

 arguments in favour of the popular mode of understanding the structure of 

 Reseda are : 



1st. That the presence and appearance of the hypogynous disk, the 

 anomalous structure of the petals, and the sestivation of the flower, all occur 

 in a greater or less degree in Capparidese, and have been found united in no 

 other family of plants ; and, 



2d. That the appendages (which I consider abortive stamens), being 

 formed before the part upon which they rest (and which I have called 

 calyx), are consequently to be referred to the corolla rather than the 

 stamens: this, at least, is how I understand the chief argument employed by 

 Mr. Brown. I hope I do not misunderstand, 



3d. That the processes of the supposed petals are analogous to those 

 of Dianthus, Lychnis, and Silene. 



To the first of these arguments I reply, that, without meaning in the 

 slightest degree to doubt the accuracy of Mr. Brown's observations, which 

 I know are beyond question, I have not been able to discover any Cappari- 

 deous plants which are in my judgment analogous in the conformation of 

 their parts to Reseda ; and that, even presuming appearances of analogy 

 to exist more unequivocally than Mr. Brown states that they do, such a 

 fact would not by itself shake the evidence I have produced to the contrary. 

 To this I may add, that analogical evidence in support of my position, fully 

 as powerful as that said to exist against it in Capparidese, is furnished by 

 Datisca, a genus I think evidently very near Reseda, which is unquestion- 

 ably apetalous, and of which the calyx of the female flowers may without 

 difficulty be compared with that of Reseda, except that it is adherent to the 

 ovarium. 



To the second objection it may be answered, that in organs of so anomu- 



