Bactris,\ eqtjatorial-ameeicak palms. 155 



Amazonuni 



Andium 



forma 



lineares, acumine brevlore et magis abrupto, costa media ssepe per 

 totam longitudinem aculeolata setosave j aculeis petioli praelongis 

 ssepe 3-pollicaribiis. 

 J'7ore5 utriuscLue sexus adsunt. $ jlavi'^ sepalaS perbrevia inrequalia 

 acuta J corolla multo longior cartilaginea trigona profunde tripartita, 

 laciniis ovatis subrepandis ; stainiiia 6 perigyna?, antlieiis vorsatilibus. 

 FL ? : calyx globoso-urceolaris, ore constricto subrepando ; corolla 

 brevier inclusa platystoma extus lanata. 



Desmokcus, Mart. 



Of this genus, whicli in its sarmentose stems and prickly cir- 

 rhate leaves is the only western analogue of the eastern Calami^ 

 although so widely diverse in its other characters, T seem to have 

 gathered but two species ; but Desmond so frequently and preju- 

 dicially crossed my path, that I must have left many other forms 

 ungathered. One of the two seems to be the D, macracanthos^ de- 

 scribed by Martins, and well figured by Wallace (plate 27). But 

 besides this species, Martins describes seven others, whereof four 

 are proper to the Amazon valley, one of the four having been 

 found by Peppig in Eastern Peru (Maynas). 



Karsten merges Martius's genera Desmoncu^ and Guilielmia in 

 the Bactris of Jacquin ; but the first of these is certainly as well- 

 marked and distinct a group of species as almost any recognized 

 genus of palms. Its very obvious differences from Bactris are : — the 

 scandent habit ; the opposite pinnte, whereof the lower are so nar- 

 rowed at the base as to be spuriously petiolate and almost exactly 

 lance-shaped, and the upper 6 or 8 pairs are transformed into 

 straight tough subulate spines so reflexed at an acute angle as to 

 catch firm hold of any soft substance that brushes against them ; 

 tlie petiole consisting almost entirely of sheath (the rhachis being 

 pinnate almost to the base, where it is abruptly decurved from 

 the stem) and tlie sheath prolonged upwards intoanocreaS inches 

 long ; whereas in the Bactridcs the sheath has either only a ligule 

 at the mouth, or it extends into a very short oblique ocrea, and, 



besides the sheath, there is a proper petiole of from 1 to 2 feet 

 long. 



Besides these differences, no Desmoncus has leaves ciliated 

 with aculeoli; while every Bactris has them so, if not all along 

 the margin, at least in a pencil at the point of the leaf. 



% 



