186 



This again is antedated by 24 years by the name Cordyline, 

 which Adanson (Fam. des Plantes, vol, 2, pp. 54 and 543) estab- 

 lished in 1763 upon S. lanuginosa, Willd. and S. zeylanica, 

 :Willd. So that if priority of generic name is strictly adhered 

 to, all the species here enumerated should be placed under the 

 name Cordyline, concerning which and other matters relating 

 to Sansevieria, see the Keiv Bulletin, 1914, p. 273. But in 

 deference to the ruling of the Botanical Congress at Vienna in 



inn:: xi,^ _•_ " c • ■■ • ^ o . 



h 



ma 



would only result in very great inconvenience and confusion. 

 In Bentham and Hooker, Genera Plantaruvi, vol, 3, p. 679,. 



am m the order Haemod 



placed, is most unaccountable and can only be understood upon 

 the assumption that he entirely overlooked the very obvious 

 affinity of Sansevieria to Dracaena, which is so pronounced that 

 several species of Dracaena (one even in recent years) have been 

 described as species of Sansevieria. Engler in his Pflanzen- 

 familien, vol. 2, part 5, p. 84, has rightly transferred Sansevieria 

 together with Opliiopogon, Liriope and Peliosanthes to Liliaceae, 



om 



so that the very close affinity of these two genera would appear 

 to have been overlooked; a discussion of their relationship will 

 be found in the Keio Bulletin, 1914, p. 273. 



The genus Sansevieria is chiefly confined to Africa and the 

 islands near its coast, and Arabia. Nearly all the species are 

 congregated in the tropical area of the continent, as only 5, if 

 my identification of S. angustiflora (p. 248) is correct, occur in 

 bouth Africa, and none are found in the northern extratropical 

 region. Only 4 species are definitely known to inhabit any other 

 region, and these are natives of Ceylon, India, Burma, and 

 perhaps China. 



L ■ 



As is well known, many species of this genus are largely culti- 

 vated in various parts of the Tropics for their valuable fibre. 

 Which IS of excellent quality in almost all, although varying with 

 the species. Unfortunately it is impossible at present to give 

 particulars cpncernmg the quality and value of the fibre derived 

 trom the various species, for two reasons. In the first place, most 

 of the fibres examined in our museums are labelled as being 

 derived from S. guineensis, S, sulcata, and S, zeylanica. From 

 the evidence accumulated during the progress of this work, I 

 have every reason to believe that neither of these names belong 

 to the fibres to which they are attached, unless it may be that 

 some samples labelled S. zeylanica were actually obtained from 

 the wild plant of Ceylon, since the Indian plant is distinct, and 

 ali living cultivated plants so named prove to belong to two or 

 three other species, whilst at least four distinct species have been 

 seen m the living state cultivated under the name S. guineensis 

 and three under that of S. sulcata, neither of them' correctly 

 named. In the second place, the fibres are not accompanied by 



