It has been already remarked, that there is nothing in the structure of the column in Rafflesia to enable us to determine the position of the ovarium in the female flower; but that from another consideration there seems a some- what greater probability of its being superior. If, however, it were even inferior, the objection to the affinity in question would not be insuperable, the relationship of Homalinæ to Passifloreæ being admitted.
If Napoleona or Belvisia be really allied to Passifloreæ,
brevissimæ geiiitalium inserta. Styli 5. Stigmata peltata, Capsula inflata, quinquevalvis. Semina axibus valvularum inserta.
Frutices (forsan decumbentes). Folia alterna simplicia suhdentata, stipulis lateralibus (utrinque solitariis geminisve) distinctis, callosis. Flores axillares subsolitarii, pedunculis, quandoque brevissimis, basi bracteolatis. Urceolus abbreviatus, ore denticulato. Filamenta simplici serie, viginti circiter. Antheræ incumbentes, lineares. Capsula chartacea. Semina axibus filiformibus valvularum subsimplici serie inserta, pedicellata, punctata, omnino Passifloræ.
Patria. Africa æquinoctialis.
1. S. pubescens, ramis tomentosis, foliis oblongo-ovatis basi obtusis: adultis pube rara conspersis, urceolo barbato.
Smeathmannia pubcscens. Solander l.c.
Loc. Nat. Guinea, prope Sierra Leone, Smeathman, Afzelius.
2. S. lævigata, ramis glabris, foliis oblongis ovatisve basi acutis: adultis glaberrimis utrinque nitidis, urceolo imberbi inciso.
Smeathmannia lævigata. Soland, l. c.
Loc. Nat. Guinea, prope Sierra Leone, Smeathman, Afzelius, Purdie.
3. S. media., ramis glabris, foliis obovato-oblongis basi obtusis: adultis utrinque glabris subopacis.
Loc. Nat. Guinea, prope Sierra Leone, Smeathman.
Forsan varietas S. lævigatæ.
The affinity of Smeathmannia to Paropsia of M. du Petit Thouars will probably be admitted without hesitation; and its exact agreement in fruit in every important point, both with this genus and with Modecca, seems to leave no doubt of its belonging to Passifloreæ, with which it agrees in habit even better than Paropsia, and certainly much more nearly than Malesherbia, considered by M. de Jussieu (in Flor. Peruv. iii, p. xix) as belonging to the same family.
Smeathmannia differs then from the other genera of Passifloræ solely in its greater number of stamina, which, however, may not be really indefinite; and an approach to this structure is already known to exist in an unpublished genus (Thompsonia) discovered in Madagascar by Mr. Thompson, of which the habit is entirely that of Deidamia, and whose stamina are equal in number to the divisions of both series of the perianthium.
But from Smeathmannia the transition is easy to Ryania, which differs chiefly in its still greater number of stamina, in the want of petals or inner series of perianthium, in the single style being only slightly divided, and in the form of its placentæ.
And Ryania, although it has a superior ovarium, may even be supposed to be related to Asteranthos and Belvisia, if the fruit of these two genera should prove to be unilocular with several parietal placentæ.
