ANONAGEJJJ. 193 



ovules, and becomes a shortly stipitate berry,' quite analogous to 

 that of an Asimina. The Uvarias of this section are shrubs with 

 alternate leaves and axillary flowers, solitary or united into few- 

 flowered cymes. Their peduncles are pretty long and slender, often 

 with a bract about half way up. We may add that several Porcdia 

 flowers that we have dissected w^ere becoming male by the more or 

 less complete abortion of the gynseceum. ^Mlen this disappears 

 entirely we find only stamens, inserted up to the very centre of the 

 receptacle, which is even less convex than in the hermaphrodite 

 flowers. We may then define Porcelia as consisting of American 

 species"^ of Uvaria with hermaphrodite or polygamous flowers. 



Nor is there any great generic difl:erence between the Uvarias and 

 Sapraiitlius mcaraguensis^ a plant w^hose enormous flower^ possesses 

 a trimerous imbricate calyx, and six large, equal, flattened, mem- 

 branous petals, forming a double imbricate corolla. The carpels 

 and stamens are indefinite, and formed exactly as in Asimina and 

 Porcelia ; so too are the fruits. This section of the genus Uvaria is 

 as yet only represented by one small tree, whose leaves are coated 

 with a velvety down, like the twigs and peduncles. The solitary, 

 leaf-opposed flower' terminates a peduncle which bears a leafy bract, 

 and is distinguished by its foetid odour, and its dull violet-blue colour. 



In the Uvarias properly so called, as well as in those of the section 

 Asimina, the inner petals are of the same size as the outer ones, or 

 a little smaller, or rarely somewdiat more developed. This last rela- 

 tion is that found, though but little marked, in two types referred by 

 most authors to a very diff'erent group" of this order, namely, Mareu- 



' This beny, whose external confignralion is B. H., Gen., 956 (Porcelia). — II. I5x., Adan- 



exactly represented in the work of Kuiz k Payox, sonia, viii. 303. 



is closely analogous to that of Ucaria. The seeds ^ The petals are " from four to six inches 



are separated from each other by a thin soft lon^-." 



prolongation of the endocarp. They are flattened ^ They are tlius represented by tlie author 



and oval, and the arillary thickening of the outer (Seemaux). who describes them, however, as 



coat is hardly indicated around the point of at- axillnry. The foliage of this phmt appears very 



taclmient, even less so than in Asimina triloba niucli like that of A. triloba. Aci'ording to tlie 



DrN. description, the imbrication of the corolhi appears 



- Only with some doubt have we referred to to be more marked than in Asimina, but this 



this group, our Ucaria Mahniana [Adansonia, character is one that may vary greatly in one and 



viii. 347, n. 11), an American species, whose the same genus, as we shall see below, 

 fruits (which have alone been examined) are very " Tliat of the MitrepfwretF, of which the 



nearly those of Porce/ia and ^«j»u«a, but whose corolla, often characteristic, is thus defined by 



seeds arc regularly arranged in two parallel rows. BeNTRAM & Hookek (Oen., 21) : " Pttala vol- 



The flowers are as yet imknown. rata, exteriora aperla, inter iora circa genitalia 



^ Seemann, Journ. of Bot- iv. 369, t. liv. — ereclo-couniveniia v. connala." 



VOL. I. O 



