TLOAVEB IN THE GENUS NAPOLEOKA, ETC. 503 



necessary to comment are the following : — The petioles are de- 

 scribed as not thickened (hand crassis). The anthers are said to be 

 ten, each one-celled ; while the fruit, of which good specimens still 

 exist in the Museum at Kew, is described as " bacca corticosa, 

 magnitudine et facie fructus Punicse, cortice extus rubescente 

 punctulis albis crebre consperso, septis pulposis in spec, nostris 

 exsiceatis et semicollapsis et cum integumento seminum conglu- 

 tinatis." Mr. Bentham, in commentiag on this description, 

 merely gives his opinion wath some doubt that the iV. Heudelotii 

 and i\r. Vogelii are not really specifically distinct from the original 

 -ZV^. imperialism 



Such, then, are the main structural points as recorded or figured 

 by various observers, or as observed by myself. I have endea- 

 voured to show how far the discrepancies between the various 

 statements may be explained or reconciled ; and I may now add, 

 in all deference, that, in my opinion, they can hardly be taken to 

 furnish points for specific distinction, seeing that they are either 

 too inconstant and variable, or, as in the case of the blue colour of 

 which Palisot de Beauvois speaks, as well as in that of the two- 

 celled anthers mentioned by Decaisne and Lindley, errors of 

 observation. 



The points of distinction cited by Decaisne (as follows : 



1. -ZV". imperialiSy flowers in groups of three, blue ; 



2. N, Seudelotii, flowers solitary, reddish-purple ; 



3. N. Whitjteldii, flowers solitary, red and yellow \ 



4. iV^. Vogelii, flowers solitary, red and yellow) 

 are obviously insufficient, as being inconstant and variable, and 

 not borne out by more complete examination. The only points 

 that remain are the form and size of the leaves, the size of the 

 flower, and the shape of the lobes of the corona ; but these appear 

 to me to indicate varieties rather than species. 



As to the morphology of the flower I have little to off'er. I 

 should regard the calyx as consisting of five sepals, the glands 

 indicating: the base of the abortive lamina) as in the true leaf. With 

 the lobes of the calyx alternate the lobes of the corolla, the ribs 

 of which latter I consider equivalent to the rays of the first 

 and second lobes of the corona, but tied together by intervening 

 cellular tissue- The segments of the corona or staminodes are 

 adherent to a certain distance to the stamens, which latter I 

 should consider "compound" or 5-lobed, some of the lobes 

 antheriferous, the others sterile. The disk is clearly a prolonga- 



lilNN. PROG. — BOTAIN'Y, TOL. X. 2 L 



