DR. R. SPRUCE ON EQUATORIAL-AMERICAN PALMS. 85 



of my own, compare them with those of other botanists, and thence 

 try to show how a more satisfactory and natural arrangement of 

 Palms may he arrived at. 



The tribes into which the order of PalmaceaB has been divided, 

 although constituting, with a few exceptions, natural groups of 

 genera, have been founded chiefly on supposed, but not very real, 

 differences in the structure of the ovary and fruit. The first 

 Tribe, Arecince, is said to have an " ovarium e carpidiis 3, rarius 2, 

 primitus connatis^ 3- vel 2-loculare, rarissime carpidio unico 1-locu- 

 lare. Fructus baccatus." 



Tribe lY., Corypliincey has also usually baccate fruit ; but the 

 ovaries (carpels) are said to be '^primitus distincta,''^ 



Tribe III., Sorassince, has nearly always drupaceous fruit : "dru- 

 pa 1-3-pyrena indivisa vel lobata, interdum bacca monosperma ; " 

 and the ovary is said to be 2-3- (rarely 4-) celled. 



Of Tribe II., Lepidocaryince^ so well distinguished by its scaly 

 fruits, and Tribe V., Cocovke^ by its symmetrical triforaminate en- 

 docarp, more anon. 



Now, if we conjoin Tribes III. and lY., and eliminate there- 

 from a few genera {Geonoma, Plicenix^ &c.) to be added on to 

 Tribe I. (Arecince), we get (from the three) two tolerably natural 

 groups ; but their true characters are by no means those above 

 stated. 



§ 2. The Ovaries of all Palms consist normally qf three Carpels, 



On examining and comparing the ovaries of Palms, we shall find 

 that their normal condition is to have three carpels (" carpidia" 

 of Martins, Endlicher, &c.), four or five only by rare exception, 

 and fewer than three only by abortion. Even in Geonoma^ which 

 has the ovary usually reduced to a single uniovulate carpel, the 

 trifid style, arising from the base of the ovary on the inner side, 

 betrays the triple nature of the ovary ; and on exploring a good 

 many female flowers of any species of Geonoma, we shall be almost 

 certain to find one or more ovaries consisting of three carpels 

 united at the base to a central style, the two sterile ones either 

 persisting as minute warts at the base of the fertile carpel, or dis- 

 appearing altogether when the latter is ripe*. 



* It will be seen from this that Geonoma has been described quite hypothe- 

 tically to have "a trilocular ovary." I first satisfied myself of its true nature 

 in 18.32, when collecting at the cataracts of the Kio Negro ; and I have fully de- 

 scribed the similar structiu-c of Wettinia in the Linnean Journal for April, 



