DB. B. SPBUCE ON EQUATOEIAL-AMEEICAN PALMS. 87 



Thia structure has been considered to pertain only to the fruit 

 of the CocointBy whose black bony endocarp renders it very con- 

 spicuous ; but it seems to me to exist also in other palm-fruits, 

 which have the stigmas at the geometrical apex of the ovary and 

 fruit, but in which it is detected with greater difficulty, on account 

 of the endocarp being a membrane of extreme tenuity. 



§ 5. Tribes are insufficiently characterized hy variations in 



the Exocarpic Structure. 



If we consult and compare the characters of the genera of 

 Palms, as given by Kxmth and Endlicher, we shall see that every 

 possible modification of the exocarpic structure occurs in Arecinse ; 

 and the same thing, with a little limitation, is to be remarked of 

 Borassinae and Coryphinse. The contrast of " carpidia primitus 

 connata " in Arecinoe, and '^ carpidia primitus distincta " in Cory- 

 phina?, is found to have no existence. In Corypha itself, and in Co- 

 pernicia^ I find the carpels from the first combined to a greater or 

 less height. If in such Coryphinous genera as ChamcBrops^ Thri- 

 naXj and Hhapis the carpels are distinct to the very base, so also 

 are tbey in the Arecinous genera Geonomaj Wettinia^ Nunnez- 

 haria^ &c. The exocarpic character serves, in fact, to unite, not 

 to separate, the tribes Arecinse, Coryphinse, and Borassinae. To 

 distinguish them, if distinction there be, we must seek elsewhere. 



§ 6. Homologies of the Pericarp of Palm fruits. 



I must here interrupt my c[uest to remark that the nomencla- 

 ture of the fruits of Palms requires to be put on a firmer basis ; 

 for in a nmltitude of cases what one author has called a " drupe " 

 another calls a " berry ; " and if it be admitted that there may be 

 both dry drupes and dry one-seeded berries, the distinction will 

 be reduced to depend on the nature of the endocarp, not on its 

 presence or absence ; for I believe it is always present in some 

 shape or other. The number of separate (or separable) envelopes 

 is by no means the same in the pericarps of all palms, some having 

 apparently only two, others three, and many four, each of which 

 may consist of several laminae that are themselves sometimes easy 

 to separate, besides the envelopes of the seeds, or " nucleus," as it 



has been called. 



In the Tagua Palm (Aftalea Ilumholdtiana, sp. n.) the envelopes 

 of the drupe are :— (1) a thin brown skin, or epicarp, clad with 

 deciduous lepra ; (2) several separable fibrous lamina) forming a 



