cook: ivory palms in Panama 139 



phas is really a palm, tho there is no adequate reason for such a 

 separation. The nearest relatives of Phytelephas are to be found 

 in other American families, such as the Cocaceae and Manicaria- 

 ceae, whose status as true palms has never been questioned.^ 



GEOGRAPHICAL DISTRIBUTION OF IVORY PALMS 



The new specimens from Panama were obtained in four differ- 

 ent localities; two near the Caribbean Coast, one near the middle 

 of the Isthmus and one near the Pacific Coast. In each case a 

 different species was secured, and all four of the species appear to 

 be different from two others previously represented in the Eco- 

 nomic Herbarium of the Department of Agriculture. In addition 

 to the places where specimens were collected Professor Pittier 

 heard reports of the existence of ivory palms to the north of the 

 Isthmus, and this information is substantiated by Dr. A. E. 

 Heighway of Bocas Del Toro, Panama, who states that ivory 

 palms exist in small numbers in several places, most of them back 

 a few miles from the coast, along the foothills. Thus it appears 

 that the geographical distribution of the ivory palms extends 

 beyond the Isthmus, so that the family Phytelephantaceae 

 must be included in the flora of North America. 



The genus Phytelephas was based originally on two Peruvian 

 species, from the eastern slopes of the Andes. One species has 

 been described from the Pacific coast of Ecuador and another 

 from the Magdalena Valley of Columbia. Thus only four defi- 

 nitely localized species are known from South America to six in 

 the region of Panama. The existence of so many local species 

 in the narrow limits of the Isthmus makes it seem probable that 

 more careful study of the South American members of the group 

 will result in the discovery of a much larger number of specific 

 forms. Even on general biological grounds it would not be 

 expected that species with such large heavy seeds would retain 

 their continuity over wide areas of distribution. The first require- 

 ment for an adequate taxonomic treatment of the group is to 



'■ O. F. Cook, Relationships of the ivory pahns. Contributions U. S. National 

 Herbarium 13: 133. 1910. 



