264 



Pomona College Journal of Economic Botany 



O. reg'ta, and has the form of a cone with a broad base. This same character is 

 offered hy fruits I have received under the name of lioystonea horinquena by Dr. 

 Franceschi from Porto Rico; this intrusion, however, is less pronounced in the 

 fruits from n. l6()5 of Sintenis, and in those of Sintenis VII is not more pro- 

 nounced than in tlie seeds of 0. regia from Cuba. 



On the whole, O. charibaea is so closely related to O. regia that it seems 

 to me only a geographical form of that plant, or in any case nothing more than 

 a second degree species. 



Figure 113. Oreodoxa charibaea. a, longitudinal section of a fruit (from a plant 

 cultivated in the Botanical Garden at Buitenzorg) ; b, longitudinal section of a 

 seed from Sintenis No. 1605; c, longitudinal section of a seed from Sintenis No. 

 VII. Oreodoxa regia. d, female flower, and e, longitudinal section of a seed 

 (both figures from the specimen No. 8, Bot. Dep. Jamaica) ; f, longitudinal sec- 

 tion of a seed from a plant collected in S. Dominga (Taylor No. 61 in Herb. 

 BeroL). 



Mr. Cook has observed that on the roots of this palm exist tubercles, pro- 

 duced by bacterioids, analogous to those of the Leguminosae and other plants. 

 The same kind of tubercles I have observed on the roots of another palm, the 

 common Trachycarpus excelsa. 



To 0. charibaea is referrable the Genus Ignotum, Urban Symb. Antill. IV, 

 131; Sintenis n. 6512 in Herb. Krug et Urban (Berlin Herb.) collected at San 

 Andre near Usuado in Porto Rico. The specimens consist only of very juvenile 

 and tender flowering branchlets with flowers very imperfectly developed ; the 

 branchlets are entirely covered by the characteristic trichomes, which I have de- 

 scribed in 0. regia, and which has enabled me to settle the question of the iden- 

 tification of these specimens. 



