﻿t 



MR. J. MIERS ON THE LECYTHIDACE^. I75 



Bhizophoracece, whicli tends to confirm my notion that the BarringtonieiB ^Y^Yii^^Xi^ 

 towards that family rather than to the Myrtacem. 



The LecytUdacecB, as above analyzed, are here divided into 12 different genera ; and I 

 now proceed to give a distinct diagnosis of each and a description of their several species. 

 Every specimen to which access was attainable has been carefully examined ; and in 

 describing the species to which they belong, one uniform system has been adopted in the 

 arrangement of their characters. In those species described by previous authors, and 

 which I have had no opportunity of examining, the characters so given are redistributed 

 under the same order of sequence as that adopted in my own enumeration, so that their 

 mutual differences may be more easily compared. It is necessary to premise that few 

 herbarium specimens can be identified with the large fruits preserved in museums. I 

 have never attempted in any case to associate them without good authority; conse- 

 quently the fruits not duly provable appear here as separate species, thus increasing their 

 number to more than circumstances would otherwise warrant ; but it is better to suffer 

 this disadvantage than to give a false character to any species. The specific features 

 afforded by the fruits are excellent, and quite equal in value to those furnished by tlie 

 plants and flowers. No difficulty whatever has been found, in the absence of the fruit, 

 to determine with tolerable accuracy, from the structure, of the ovary in the biul, tlio 

 genus to which any specimen belongs— a result which can hardly be said to have been 

 attainable before. This is a great point gained in our knowledge of .the family. The 

 characters, especially those of the ovary, are therefore carefully notqd in the several 

 species, as they serve to justify the position to which the different. specimens are assigned. 



1. GusTAviA (Plate XXXIII. a). 



This beautiful genus, one of the oldest of the family, has been considered by most 

 botanists to belong to the Barring toniecB, or has been regarded as an anomalous 

 member of the Myrtacece ; but my analysis demonstrates that it is truly Lecy ihtdaceon a. 

 The most prominent rank is here given to it, as the typical genus of the family, because, 

 from its more regular development, it affords a key to a better comprehension of tho 

 smgular structure found in the other genera. It differs little, in fact, from Courotipita 

 and Cariniana except in the regularity of its development, all the parts being quite 

 similar in their nature and position. Grias more nearly approaches it in the regularity 

 of the parts of the androphorum, although these are more stunted in growth. 



GusTAViA, Linn 



Pirigi 

 Calyx 



Marcg 



,' 



4-6 



magna 



planus. 



firuibus 



annul 



magnum 



petalorum 



latum, et deinde cum petalis deciduum, imo usque ad medium, lu 

 cyatlium latum, brevem, camosulum, e laminis plurimis extus gradatira majoribus arete agglutinatis 

 formatum, laminarum marginibus liberis, et undique in lacinias, seu appendiculan, numcrosas, subu- 

 lato-loriformes fimbriatim divisis, serie interiore brevi, reliquis imbricatim sensim longioribus, om- 



2a2 



