400 NATURAL HISTORY OF PLANTS. 



filaments are free to a much larger extent, sometimes even nearly 

 to the base. Their number may rise to eight or twelve and perhaps 

 even more.^ In conclusion, the genus Quapoya, thus limited,^ is 

 developed almost parallel to the genus Clusia, in which we have 

 seen variations of the androecium still more numerous, both as to 

 the number of stamens and the form of the anthers. It comprises 

 some fifteen species,^ belonging entirely to tropical America ; the 

 habit and organs of vegetation are those of Olusia, but the flowers 

 are ordinarily much smaller. 



Havetia * has dioecious flowers, and the leaves are nearly the same 

 as those of the preceding genera. The ovary, surrounded by an 

 hypogynous disk,^ has generally in each of its four cells two ^ 

 descending ovules, with micropyle interior and superior, and raphe 

 ventral and sublateral.'^ But the male flowers, ordinarily tetramerous, 

 with four imbricate petals, are remarkable for the androecium, com- 

 posed of four large alternipetalous stamens ; each of which has the 

 form of a thick quarter of a sphere, and bears above and without 

 three circular and yalvicide cells. The only species of Havetia^ 

 known is a Columbian tree having otherwise the foliage, habit, and 

 inflorescence of Quapoya. 



Beside the preceding genera under the name Clusiella ^ has been 

 placed, not without some doubt, a Columbian shrub having penta- 

 merous dioecious flowers. In the females, alone known, there are 

 contorted petals, and an ovary with five multiovulate cells, sur- 

 rounded at the base by a cupule formed of a large number of sterile 

 stamens, short and closely united. The flowers, small and collected 



1 To 20 in the sect. Hemiquapoya (Pl. et ~-'Q.'H..Gen.in,-D..'i.— R.B^. Payer Fam. Nat. 

 Tri. loc. cit. 288), and if, as the same authors 270. 



suppose, Arrudea? bicolor Bemth. belongs to ^ It has been supposed to be formed by the 



this genus, it would be the richest representa- union of four staminodes in a sort of cupule. 



tive in stamens since the latter number about * It may, it is said, have four, then two 



forty. inferior ascending (B. H.). 



2 Sect. 6 : 1. Euquapoya {Rengifa) ; 2. Eave- 7 The raphe becomes dorsal or nearly so at 

 tiopsis ; 3. (Edematopus ; 4. Balboa ; 5. Renggeria ; adult age in Pilosperma (Pl. et Tri. Ann. Sc. 

 6. Hemiquapoya. Nat. ser. 4, xiv. 243. — B. H, Gen. 171, n. 4), a 



3 Mart. Nov. Gen. et Sp. iii. 166. t. 297, f. iii. Columbian tree which has the characters of 

 (JTaw^ta).— Benth. Hook. Lond. Journ. ii. 369 vegetation of Havetia, and of which the tetra- 

 {Ilavetia), Eew Gard. Misc. iii, 146 {Arrudea ?). merous female flowers only are known ; but we 

 — PcEPP. etENDL. Nov. Gen. et Sp. iii. 11, t. 209 A know not if the raphe may not primarily be 

 {Havetia). — Walp. Rep. i. 493 {Rengifa); ii. equally ventral. Is the uril (?) of the seed, as 

 810 {Havetia); Ann. vii. 343 {Rengifa), 344 supposed, distinct in origin from thoi oi Have- 

 (^Havetiopsis, (Edematopus), 345 {Balboa). tia ? 



•» H. B. K. Nov. Gen. et Sp. v. 203, t. 462.— « ff laurifolia H. B. K. loc. cit. (not alior.). 



Spach, Suit, a Bufon, v. 305.— Endl. Gen. n. » Pl. et Tri. Ann. Sc. Nat. ser. 4, xiv. 253.— 



5435.— Tri. et Pl. Ann. Sc. Nat. ser. 4, xiv. 245. B. H. Gen. 172, n. 7. 



