﻿174 MR. J. MIERS ON THE LECYTHIDACEiE. 



generally large, each often suspended by a fleshy funicle nearly as large as itself, often 

 imbedded in pulp, at other times dry, erect and sessile in the base of the fruit, winged 

 or bare, often costate, all with an exalbuminous embryo, sometimes of singular develop- 

 ment. We have here a sum of remarkable characters, in reference to which we find no 



It 



parallel whatever in the Myrtacece. 



On the other hand we see in the Myrtacece^ properly restricted, opposite leaves, T\dth 

 punctate glands, often concealed within a thick parenchyma, but generally apparent, pel- 

 lucid, and of aromatic flavour : the flowers, generally small, 4- or 5- rarely 6-merous, have 

 an inferior ovary, with 4, 5, 6, or more cells, with many horizontal or ascending ovules ; 

 a circular annular disk agglutinated to the free tubular portion of the calyx, and there- 

 fore unmistakably perigynous, which bears the petals upon its outer margin, and is 

 charged all over wdth numerous stamens in several series, except when, rarely, they are 

 imiserial ; the filaments are very long and slender, coiled and replicated in aestivation : 

 the disposition of the calyx, charged with its staminiferous disk, is so peculiar that 

 it has received the name of an hypanthium : the fruit, in the first tribes, is generally 

 capsular, but drupaceous or baccate in the Myrtea : all are plurilocular, or by abortion 

 rendered 1-celled, with bare fleshy erect exalbuminous seeds ^ Compared with these 

 trenchant features, no correlative characters are to be found in LecytJiidacece. 



Under such circumstances it appears to me injudicious to agglomerate into a single 

 natural order groups so utterly distinct as the Myrtacece, LecytJiidacece^ Barringtoniaceoe 

 and Napoleonece : a family so congregated can hardly be said to possess a single positive 

 peculiarity, because under the fusion of so many opposing characters, each of the ordinal 

 features necessarily becomes negative, and, according to my view, tends to destroy the 

 uniformity and utility of any system of arrangement based on such a principle. 



On a future occasion I will proceed to detail the structure of the BarrmgtmiacecB, 

 and will endeavour to show by evidence the affinity they bear to the above and to other 

 families. I propose also to demonstrate that the Napoleonece have no relationship with 

 the above order or with Myrtacece, and that the 2 genera composing it have no mutual 

 affinity, and belong to other different orders. I am also prepared to deal with the 

 many dubious genera referred to the group of the Lecythidaceae: Couponi, AxMd, 

 forms a singular genus {Cupirana) of the Apocynacece; Cupheanthus, Seem., certainly 

 belongs to the Lythracece; Foetidia, Lam., appears to me to belong to the same family; 

 Calostemma, Benth., seems referable to the Fomece in Rosacece ; Tropiera off"ers a near 

 affinity with Glossopetalum in Celastracece. And I have also to describe and figure a new 

 genus Sarmena, founded on a plant from New Caledonia, noticed by Dr. Seemann in his 

 PL Vit. as a specimen of Barringtonia speciosa ; this forms an interesting genus of 



w 



* Dr. Berg makes a distinct tribe of his genus Feijoa (Orthostemon) becanse of its albuminous seeds (Mart. H. Br. 

 /. c. pp. 467 & 615). The tiro species on which it is established do not appear to differ much from others growing io 

 the same locality, referred to his genus Myrdeugenla. He seems to have had some misgiving on the subject, because 

 he neither describes the albumen in the copious details of the text above cited, nor has he figured it in plate 54, 

 where the ripe fruit is shown. So loose a statement cannot be received as evidence of a fact of such importance, 

 wliich would form a unique exception to the universal structure of the whole family. It will be remembered that 

 a very analogous error was made in the same work in the case of Maytenus (see linn. Trans, vol. xxvii. p. 326). 



* 



t 



