RELATIONSHIPS OF THE IVORY PALMS. 



By O. F. Cook. 



HISTORICAL OPINIONS. 



The plants that yield the South American vegetable ivory belong 

 to the genus Phytelephas. Though popularly reckoned as palms 

 their claims to this distinction have been rejected by eminent botani- 

 cal authorities. It may be that this is one of the cases where general 

 appearances gave a more correct indication than technical distinc- 

 tions formulated by specialists. Some of the characters of Phytele- 

 phas have been wrongly stated, and thus made to appear more differ- 

 ent from those of other palms than they really are, and some of the 

 analogous specializations of other 1 palms have been overlooked. 



Martius, the most celebrated student of the palms, deferred to the 

 popular impression to the extent of including Phytelephas in his 

 monograph, but held that it should be classified as an independent 

 family intermediate between the palms, the screw pines, and the 

 aroids. Later writers, such as Hooker. Seemann, Spruce, and Drude, 

 have followed Martins in the general policy of denying that Phytel- 

 ephas is a true palm, though differing in their views of its relations 

 with other families. 



Kunth. in describing the South American plants collected by Hum- 

 boldt, placed Phytelephas in a second section of the Typhinae, be- 

 tween the aroids and the grasses, remote from the true palms. The 

 only character given by Kunth that would distinguish Phytelephas 

 from other palms is the statement that it has a single spathe, "spatha 

 monophylla" evidently an error in fact, for other observers have 

 reported two or three spathes. Karsten afterwards described in 

 greater detail the Phytelephas of the same region where Humboldt's 

 plant was found, the upper valley of the Magdalena River of 

 Columbia. 



Seemann also believed that the affinities of Phytelephas lay with 

 the Old World screw pines (Pandanaceae). He looked upon the 

 Cyclanthaceae as the other American representatives of the Pandanus 

 series, and allowed them to stand between the true palms and the 

 " Phytelephantheae." The following statement indicates the reasons 

 for this arrangement: 



" In habit, the Phytelephas macrocarpa resembles the corozo Colo- 

 rado {Elan melanococca Gaertn.) ; so much so, indeed, that at fir>t 



133 



