158 DR. R. SPRUCE OX [^Astrocavyum. 



lip iuto about six lacinise, that spread back like a stellate corolla, 

 and finally allow tlie endocarp to fall out- It was doubtless this 

 peculiarity which led Meyer to give the name ^' Asfrocaryum^^ to 

 his original species (A. aciileatum, Mey. Esseq. 266), which, 

 along with A. Miinhaca and gi/nacanthuQu, are probably the only 

 true species of the genus — the other species included under 

 Astrocaryum by Martins all varying considerably from the type, 



(so far as I have seen) 



pericarp 



The Astrocarya are described by Martins and others as having 

 a simple spathe ; but those which I have been able to examine 

 have certainly two spathes, whereof the outer and thinner often 

 falls away at an early stage, which perhaps accounts for its having 



been overlooked. 



? 



Martins 



intemus basi in annul um membranaceum ele- 



vatus. 



»♦ 



A, humiley and probably some other species included in Astro- 

 caryum by Martins and Kunth, correspond precisely to Karsten's 

 " Guilielmia " in the two spathes, the gamophyllous flowers of 

 both sexes, and the andro3ceum of the $ flowers adnate to the co- 

 rolla. But many botanists will choose to keep the original ^'Gui- 



(of Martins) 



cd 



(in any case) 



from the species to which its author applied it, and appropriated 

 to a totally distinct group. Karsten's genus " Marara *' differs 

 from his " Ouilielmia " in having the calyx of both sexes, ias well 

 as the corolla of the males, triphyllous, but agrees with it in the 

 adnate androeceum and other characters. But unless there be 

 some degree of imbrication as well as complete separation of the 

 petals, the latter character is of little generic value, as is proved 

 by some species of Geonoma having the petals free to the base, 

 while their near allies have them united into a gamopetalous 

 corolla. 



Until, therefore, the whole tribe of Prickly Cocoinese shall have 

 been revised from ample materials, I prefer to continue to place 

 the two species below described in Astrocaryum^ where Martins 

 placed them, although neither of them agrees perfectly with 

 his character of that genus, nor are they themselves indisputably 

 congeneric. 



