77 
Flora of Narcondam and Barren Island. 
122. Pholidota imbricata Lindl. Barren Island (54), inside crater. 
India, Bui-ma, Malaya. 
XXXVIII. SCITAMINE^. —. 
123. Musa sapientum Linn. The Flayitain. Narcondam (104), a large 
grove behind the Coco-nut trees at Coco Bay. 
Cosmopolitan in the tropics, cultivated. 
No doubt deliberately introduced for the benefit of possibly ship-wrecked 
mariners, though it is not quite clear who planted it; probably (see under Cocos 
nucifera) it has been introduced from the Andamans, and perhaps dates from 1866. 
XXXIX. DIOSCOREACE^. xxx. 
124. Dioscorea sativa Linn. Narcondam (105) Barren Island (55). 
India, Burma, Malaya ; Australia. 
125. Dioscorea glabra Roxh. Barren Island (56); common. 
India, Burma, Malaya. 
XL. LILIACEAE. xxxi. 
126. Drac^.na ANGUSTIFOLIA Roxb. Narcondam (106), Anchorage Bay. 
Indo-China, Malaya, Australia. 
127. Gloriosa superba Linn. Barren Island (57), E. coast near sea. 
Africaj India; Indo-China, Malaya. 
XLi. COMMELINACEiE. — . 
128. PoLLiA Aclisia Hassk. Narcondam (107), very abundant on 
slopes overlooking south end of Anchorage Bay. 
Eastern Himalaya, Indo-China, Malaya. 
XLii. PALME.^. XXXII. 
129. Cartota mitis Lour. (G. sololifera Wall.) Narcondam (108). 
Indo-China, Malaya. 
130. Cocos NUCIFERA Linn. Narcondam (109), many at Coco Bay, 
a few at Anchorage Bay, one, not yet bearing, at E. Bay; Barren Island 
(58), thirteen trees counted from the offing, behind the Pandanus fence 
at Anchorage Bay; none seen elsewhere. 
India; Malaya ; Polynesia; America. 
The introduction of this tree into these islands is a question of some interest. 
The tree at E. Bay, Narcondam, has no doubt been produced from a nut washed 
round from Coco Bay; in all likelihood the trees at Anchorage Bay have been 
derived from the same source. The trees at Coco Bay itself may have origi¬ 
nated from nuts brought from the Coco Group by a surface-current sweeping from 
the Sea of Bengal, through the Preparis Channels, from N.-E. to S.-W. across the 
Andaman Sea; but as they are associated, where they occur, with a grove of 
Musa sapientum (which must have been deliberately introduced), it is not unreason¬ 
able to suppose that the two species were introduced together. 
291 
