04 USEFUL PLANTS OF (JUAM. 



tivatcd, but VGi\y little bus been done to fix tbc species and varieties 

 and to compare tbose growiiif;' in ditl'erent parts of the world. Yams 

 are dioecious, and the tlowers of man}' recoj^iiized varieties are imper- 

 fectly known. In some (;ases the flowers of but a single sex have been 

 described; in others the fruit has never been observed, and in others 

 only the tul)ers are known. Sir Joseph Hooker," who has done much 

 to straighten out the Indian species, writes as follows: 



The species of Dioscorea are in a state of indescribable confusion, and I can not 

 hope to have escaped errors in the determination and delimitation of the Indian 

 ones, to which I have devoted much labor. The Roxburghian food-yielding species 

 arc for the most part indeterminable, and, except through a knowledge of them as 

 cultivated in India, they can not be understood. No doubt some of the species 

 described by me have other earlier names in the Malayan flora than I have given; 

 but the Malayan species are even more loosely described than the Indian. The 

 Wallichian collection is very complete, but the species are often mixed. 



What has been said of the Indian yams applies also to those of the 

 Pacific islands, and is also true of the many varieties of Musa and 

 Artocarpus. Nearly every collector gives a list of named varieties of 

 Dioscorea, Musa, and Artocarpus in the vernacular of the various 

 localities visited, but scarcely any attempt has been made to fix these 

 varieties and to bring together the various kinds from different local- 

 ities for comparison. These must be studied in the countries where 

 they are found and should be represented in collections not only b}'^ 

 series of botanical specimens of the flowers, fruit, leaves, and roots 

 (in alcohol, when necessary), but by photographs of the fresh plants, 

 including representations of the flowers, fruits, tubers, etc., of natural 

 size or according to some definite scale of reduction or enlargement. In 

 this way only will it be possible to bring together and compare species 

 and varieties from India, Australia, the Malayan and Pacific islands, 

 Africa, and America. 



SCREW PINES. 



The Pandanaceae are known no ]>ettcr than the yams. Some of them 

 are propagated asexually for the sake of their textile leaves, and nmch 

 confusion exists among the species. Very few have been described. 

 Warburg has done much to delimit the species and varieties and clear 

 up questions of synonymy, but there remains much more to be done. 

 In his monograph of the Pandanaceae* Warburg mentions only one 

 species, Pandanus duhius Spreng. {Ilomhronia edulls Gaudich.), as 

 occurring in the Marianne Islands, and does not refer to the textile 

 species with glaucous leaves (the aggak of the natives), which has been 

 cultivated in Guam from prehistoric times (PI. VII), nor the fragrant- 

 fruited species with bright green leaves {Jcafo)^ which is one of the 

 most common plants of the island (PI. LX). As only one sex of the 



a Hooker, Flora British India, vol. 6, pp. 288-289, 1892. 



b Warburg, Pandanaceae, in Engler, Pfianzenreich, vol. 4, p. 9, 1900. 



