Miscellaneous. 143 



of corresponding, as in Orchidece, with the anterior segment of the outer 

 series of the perianthium, is placed within the posterior segment of the 

 inner series, the two remaining barren stamina of the same series 

 being the epigynous glands or filaments existing in all the genera of this 

 order except Castus ; s while the outer series of stamina, very differ- 

 ently modified, form the innermost, or supplementary series of the 

 perianthium. 



This view of the origin of that series was many years ago communi- 

 cated to me in conversation by the celebrated Correa de Serra ; but 

 was first, I believe, published in 1836 by Professor Lestiboudois in a 

 Memoir 6 in which the correctness of the opinion held, namely, that 

 Scitaminece and Cannece possess rudiments or modifications of six stamina, 

 is remarkably contrasted with the erroneous views taken, or rather 

 adopted, of the greater part of the structures adduced in support of it. 



A more accurate account of the relative position of parts was 

 given in 1828 by my ingenious friend Professor Von Martius. In 

 confirmation of the opinion, I may remark, that the cells of the ovari- 

 um, whose relation to the floral envelope appears to be very uniform 

 in Monocotyledones, are in Scitaminea opposite to the supposed petali- 

 form stamina, and to the divisions of the outermost series of perian- 

 thium. I have formerly pointed out the difference in position of the 

 antheriferous stamen in Scitaminece and that of Cannece or Marantece, 

 and have remarked that this difference is in some degree analogous to 

 that existing between Cypripedium and the other genera of Orchidea. 1 



Apostasice in its trilocular ovarium differs from all the genera of 

 Orchidece; but an analogous difference occurs in Scitaminece, in which 

 Globba is distinguished from every other genus in having its ovarium 

 unilocular, with three parietal placentae. And in both these families it 

 may be proved that the constituent parts of the compound ovarium, 

 whether unilocular or trilocular, agree in position, or in their relation 

 to the division of the perianthium. 



Lastly, Apostasies in the economy of impregnation, or the state of 

 the pollen, and the manner of its application to the stigma, probably 

 differs essentially from all Orchidece, except perhaps Cypripedium and 

 possibly Vanilla. But a similar difference, and in a degree still more 

 striking, exists between Apocinece, as I have formerly proposed to 

 limit that order, and Asclepiadece, which can only be regarded as a 

 subdivision of the same natural class. 



5 Prodr. EL Nov. Holl. i. p. 305. 



6 Memoire Sur le canna Indica et des Sur les Families des Balisiers et des Bananiers. 



7 Appendix to Denham and Clapperton's Travels, p. 243. 



8 Gen. Remarks on Bot of Terra Austr. p. 43. 



