DILLENIACEJE. 



109 



been separated, and to which some day, perhaps, it will have to be 

 restored. The calyx consists of five imbricated sepals which persist 

 and grow thick around the fruit. The petals' are imbri- 

 cated, and the stamens free and equal, or smaller as 

 they are more external. The anther cells, marginal or 

 slightly introrse, dehisce near the summit by pores 

 or ver}'- short clefts. The number of carpels varies. 

 When there are but five, as in W. hradeata. Hook. F. 

 & Thoms., they are opposite the petals. The central 

 axis uniting them touches each only along a narrow, 

 nearly vertical, line;^ whence the cells of the ovary, 

 multiovulate as in Billenia, are separated by a double 

 wall and a deep sinus (fig. 158). The carpels, the 

 walls of which are membranous, or thick and coriaceous 

 when ripe, remain indehiscent, or dehisce along the 

 inner angle to free the seeds which possess thick coats 

 covered by the fleshy aril, and copious fleshy albumen 

 with a small embryo near its apex.^ 



In certain species of Wormia the 

 innermost stamens, much longer than 

 the others, are reflexed and bent down 

 under the styles; these have been 

 erected into a genus under the name of 

 Capellia.^ In others, which there is no 

 need to separate further genericallj^ 

 the outermost stamens become sterile 

 staminodes.' 



This genus consists of trees, natives of tropical Asia, Oceania^ 



Wormia ferrvginea. 



Fig. 158. 



Transverse section of 



gynceceum. 



i 



Wormia return. 



Fig. 157. 



Stamen after 



dehiscence. 



Suit, a Buff., vli. 113.— Endl., Gen., n. 4750. 

 — B. H., Qen., 13, n. 9. — H. Bx., Adansonia, 

 vi. 281. — Lenidia Dup.-Tii., Gen. Madag., n. 57. 

 GArciCHATTD bas described {^Uranie, 476, t. 

 99), a Wormia apetala, the flowers of which he 

 describes as absolutely apetalous. We have seen 

 authentic specimens of this species, but in such a 

 condition that we were unable to say whether 

 the petals were absent before the expansion of 

 the flowers. 



^ Perhaps this line really answei-s, not to the 

 internal angle, but to the organic base of the 

 Ciirpel. In this respect, tlie gyna;ceum of Wormia 

 is no doubt comparable to that of Nigella or 

 Magnolia, in which the line of insertion of the 

 carpels is also much extended vertically. 



3 Ghiffith (Icon, posth., t. dcxUx.) hasrepre- 

 scnted the seed of his W. niffruticosa with its 

 albumen and coats slightly curved towards the 

 apex. 



* BlTjME, Bijdraj., 5. — Endi., Gen., n. 4746. 

 — Walp., Ann., iv. 34. — A. Gbat, Amer. 

 Explor. Exped., 15, t. i. 



* This occurs, but not constantly, in a new 

 species from Madagascar, which we have named 

 W.ferruginea (see Adansonia, vi. 268, vii. 343), 

 which may be distinguished at first sight by its 

 unilateral inflorescence, and the rust-coloured 

 hairs with which every part is coloured. 



« MiQ., Ann. Mus. Lugd.-Bat., i. 315, t. 9. 

 — Seem., Fl. Vitiens., 12.— BEyiii.,FL Austral., 

 i. IG. 



