COOK — RELATIONSHIPS OF THE IVORY PALMS. 135 



"fleshy'' is equally difficult to understand. In any event, no very 

 serious distinction between orders or families could be based on a 

 mere difference of consistency. The albumen is very hard in some 

 palms and relatively soft in others. 



Another South American palm, Wettinia, was at first thrown out 

 of the group by Endlicher, because it also has a simple female in- 

 florescence, suggestive of the screw pines and the cycads. The mis- 

 take regarding Wettinia was corrected by Spruce, who recognized 

 the genus as a relative of Iriartea and other closely allied South 

 American palms. It was in dealing with Wettinia that Spruce gave 

 his reason for treating Phytelephas as a distinct natural order. 



" In fact, the American plants, formerly referred to as screw pines, 

 seem to me to constitute two distinct orders, each of equal value with 

 Palmaceae and Pandanacene, viz. first, Phytelephantaceae, which are 

 (so to speak) palms with an inferior ovary; and second, Cyclantha- 

 ceae. whose inferior ovary alone separates them from Arads. Wet- 

 tinia. however, is far removed from both these: the fruits are supe- 

 rior, and though so densely crowded on the spadix as to suggest the 

 inferior concrete fruits of Phytelephas, there is no real resemblance 

 to the latter." 3 



Here again a definite difference is alleged where none in reality 

 exists. Spruce must have looked upon the conical protuberances that 

 cover the clusters of Phytelephas nuts as representing the floral 

 envelopes of "concrete fruits." The protuberances are formed in 

 reality by the splitting of an outer layer of the fruit, quite a^ in 

 other South American palms that belong to the genus Manicaria. 

 The detailed figures of Phytelephas published by Seemann and 

 Karsten show that the ovary is no more inferior than in other palms. 

 It is inclosed by a ring of staminodia, as well as by the large subulate 

 petals. Each of the several large " capitula " or "heads" that form 

 the fruit cluster of Phytelephas represents the ripened pistil of a 

 single flower, just as in Manicaria. 



Drude, in Englers Pflanzenfamilien, associates Phytelephas with 

 the Malayan genus Xipa to form a subfamily " Phytelephantinae 

 (palmae anomalae)." No characters that would call for a separa- 

 tion from the palms are included in the description of the family 

 or in that of the genus Phytelephas. The omission of the erroneous 

 statements of the older authors only makes it the more difficult to 

 understand why the custom of associating Phytelephas with an 

 Asiatic plant instead of with its American relatives should continue 

 to be followed. 



n Spruce, R., On Five New Plants from Eastern Peru. Journal of the Linnean 

 Society, vol. 3, p. 191. (1859.) 



