AN0NAGE2E. 



227 



tened blades that surmount the sort of cap thus formed, are broad, 

 tapermg towards the apex, and corrugated in the bud,' where they 

 only touch by then' edges. The whole corolla falls off in a single 

 piece. The calyx consists of three valvate leaves.' The stamens are 

 indefinite, surmounted by a truncate prolongation of the connective. 



Hexalobus (jrandifiorus. 



Fig. 279. 

 Flower. 



Fia. 280. 

 Carpel. 



The number of carpels is also indefinite, but small. ^ Each ovary con- 

 tains an indefinite number of ovules in two parallel rows,"* and is 

 surmounted by a style with two lateral papillate lobes, with tlie 

 edges rolled up.' The fruit consists of a small number of many- 

 seeded berries. Two or three species" of Hexalobus are known, 

 natives of tropical Africa ; trees or shrubs with alternate leaves and 



' They are especially covered with parallel 

 horizontal plaits, so that in the young buds of 

 H. senegalensis A. DC. (Uvaria monoiyctala 

 Rich., Guill. & Peer., Tent. Fl. Seneg., 8), 

 the apex of the petal comes very near the base. 



^ The edge is often slightly reduplicate. 



^ There are often six ; in this case each seems 

 to be superposed to a corolla lobe. 



* Bentham & Hooker state that they are 

 sometimes in one row, sometimes in two (Gew. 

 !ir>n). Hnt we have shown (Adansonia, viii. 332) 

 tliut in the species that are undoubtedly of this 

 tr^nus there are always two vertical rows of 

 Dvulcs placed back to back. 



* Each of these two lobes is a large triangular 

 hlade with its upper cdgclobod ;uid papillate, and 

 it looks as if it had been twisted up into a cornet. 



like a sheet of paper. There is, moreover, as 

 sliown in fig. 280, a terminal median lobe, 

 relatively very short and obtuse. The floral 

 receptacle is nearly plane in JI. grandijloriis, 

 and depressed and surrounded by a projecting 

 ring in II. senegalensis (see Adansonia, viii. 

 329). 



•* Rich., Gttit.l. & Pekr., Tenl. Fl. Senegamb., 

 loc. cit., t. 2. — Henth., Trans. Linn. Soc, xxiii. 

 467, t. 49.— Walp., Eep., i. 80.— OLn\ F/. 

 of Trop.Afr., 2G. We have described (Adansonia, 

 viii. 3 IS) a doubtful species of this genus. JI. 

 Irasiliensis A. S. H. & TUL. belongs to Trigi/iieia 

 (see ]i. 206) ; and perhaps, too, JI. madaga.<ica- 

 riensls A. DC. {Mem., 37, n. 2), slundd, we have 

 said, bo referred to the genus Monodora {Adan- 

 sonia, viii. 301). 



Q 2 



