VIOLACE^E. 91 



nected with the lateral ones at the very base only. The filaments 

 are continued into a membranaceous, yellowish-brown, oblong-ovate 

 and obtuse, entire, petaloid appendage, or production of the connective, 

 which much exceeds the introrsely ad n ate anther. ' The two cells of 

 the anther are narrowly oblong, longitudinally dehiscent, as in the 

 order generally ; but the apex of the anther is free, and tipped with a 

 (simple or double) vnucro (as in several Msodeiae), which in some 

 cases is conspicuous, in others nearly obsolete. Ovary globose, canes- 

 cently pubescent, one-celled, with 3 parietal rnany-ovtded placentas. 

 Style elongated, glabrous, somewhat thickened upwards, slightly curved, 

 the apex more thickened and produced on one side : the stigma small 

 and simple, terminating the lateral projection. Ovules indefinite, 

 occupying several rows on each placenta, obovoid, anatropous. Fruit 

 unknown. 



On account of the great thickness of the walls of the fertilized ova- 

 ries (from which the sepals have fallen) I have little doubt that the 

 fruit is a berry : but even if it prove to be capsular, the structure of 

 the stamens, with the deciduous sepals, will sufficiently distinguish 

 the plant from Ionidium, of which it has not the habit ; and the very 

 different corolla equally separates it from Gorynostylis, Mart. & Zucc. 

 ( Calyptrion, Gingins) . 



As the ancients garlanded the graves of their deceased friends with 

 violets, so I dedicate this new genus of Yiolaceae to the memory of 

 Alfred T. Agate, the Botanical Artist of the Expedition, who died 

 at Washington shortly after its return. The engraving is copied from 

 a drawing made by Mr. Agate, from the fresh plant, at the Feejee 

 Islands. I trust the name Agatea will be thought sufficiently distinct 

 in pronunciation both from Agathea and Agati. 



The odd and saccate petal in this plant, and probably in the whole 

 Violet Family, is truly anterior (vide Gray, Genera Floraa Am. Bor.- 

 Or. Illustrata, 1, p. 183), and does not become so by resupination, as 

 the order is characterized by Bartling, Endlicher, &c. 



Plate 7. — Agatea violaris. Fig. 1. Diagram of a flower. 2. A 

 flower, with its pedicel and bractlets. 3. Lower petal. 4. A lateral, 

 and 5, an upper petal. 6. A flower, with the calyx and corolla 

 removed. 7. The andrcecium spread open; seen from without. 8. 



