﻿MR. J. MIERS ON THE LECYTHIDACE^. 165 



(part of the dissepiments), and is furnished inside, within the margin, with a velarium 

 (a name given to the portion descending within the mouth of the shell, seen in JEschwei- 

 lera and Jugastrum) ; the coriaceous pericarp is 4-celled, cruciately divided by persistent 

 membranaceous dissepiments, which meet in the centre, without any thickening in the 

 axis : when the fruit ripens, the operculum does not fall off immediately, but remains for 

 some time, until the membranaceous dissepiments become lacerated by decay. As the ovules 

 were erect within the ovary, so we find the seeds, which fill the cells of the fruit, 

 attached to the bottom by a large basal liilum ; these are large, 1 to 3 in each cell, have 

 no funicle, are of an oblong-oval shape, erect, and, according to Dr. Spruce, in his species 

 tnrbinafa, " are enveloped in an exceedingly aromatic arillus, which the ants speedilv 

 devoured" \ This, I presume, was mucilaginous ; for I found no trace of it in the dried 

 state. Their testa is thinly coriaceous or testaceous, showing on the ventral side a 

 broadish ribbon-like band, extending from the base to the apex, free on both edges, and 

 only connected with the testa by a nerve-like line ; this band contains a cord of spiral 

 vessels, and is the main branch of the raphe ; the testa is also marked by several broad, 

 prominent, cancellated ramifications, also furnished with spiral vessels, and leaving the 

 intervening depressed areoles slightly granulated, while the branches are paler, smooth, 

 and polished : the contained nucleus, enveloped in a membranous integument, is an 

 exalbuminous embryo, tough in texture, of a dark greenish colour, of an extremely bitter 

 taste, homogeneous throughout, and ^ery different in appearance from the white, sweet, 

 amygdaloid kernel of the Sapucaia-nut ; from the concentric line, visible in a longitu- 

 dinal or in a transverse section, it seems formed of a thick exorhiza, encircling a much 

 smaller neorhiza, without the signs of any cotyledon. Aublet's Lecythis amara may 

 be considered the type of the genus, which he has figured in flower and in fruit ^ 



The genus Eschweilera (Plate XXXIV. c), originally suggested by Von Martius, but 

 afterwards confounded with Lecytliis by Berg, is here at length established upon a satis- 

 factory basis. De Candolle, upon the brief and insufficient notes of Martius, first pub- 

 Hshed it in 1828 '\ hinting at the same time that it might prove to be only a section of 

 Lecythis ; but in 1837, in order to correct this misunderstanding, Martius gave a fuller 

 outline of his genus *. Endlicher, in 1841, again complicated the matter, by absorbing 

 Eschiceilera into Lecythis, on the imperfect ground suggested by De Candolle "' ; he was 

 then evidently unacquainted with the diagnosis of Martius published four years previously. 

 All subsequent botanists followed Endlicher : even Berg, author of the beautiful mono- 

 graph of the LecytUdacece, published under the aegis of Martius, with the aid of his 

 valuable materials, so far mystified the subject that he had to modify the characters of 

 Lecythis in order to bring Eschweilera within its limits, quoting the memoranda of 

 Martius only as foot-notes. It is clear that Berg had no precise knowledge of the struc- 

 ture of either genus. Eschweilera comprises numerous species, all trees, some of great 

 magnitude, others only low trees, all with the habit and inflorescence of Lecythis. The 

 fiower has 6 small sepals, 6 much larger petals, and the usual androphorum, the hood bein 



o 



' Hook. Kew Journ. v. p. 170. » PI. Guian. ii. 716, tab. 286 & 285 a. 'J^""^^' ""' ^^^ 



* Bot. Zeit. (1837) xx. part 2, p. 89. * ^*^°- 



Plant 



XXX 



