RADACK AND OTHER ISLANDS. 147 



and that the southern groups in particular, which we 

 have not seen (Arno, Meduro, and Mille), liaving 

 richer soil and older vegetation, must produce 

 more plants, which are entirely wanting on tlie 

 northern and poorer islands. The vegetation of 

 this chain seems to have begun in the south, and 

 that man pursued its progress to the north. 



By gar, still desert and without fresh water, is 

 only visited for the purpose of catching birds, and 

 turtles. Udirick, a reef of inconsiderable circum- 

 ference, and poor in land, has only two inhabited 

 islands. In these the cocoa rises indeed above the 

 rest of the wood, but the vegetation seems poor, 

 and the bread-fruit tree rare. Tegi, near Udi- 

 rick, desert and of a scanty verdure, is scarcely 

 known by name among the people of Radack. 

 Eilu (perhaps, more correctly, Eilug) is the poorest 

 of the groups at which we landed. Udirick and 

 Eilu obtain the aroma, a plant which they have not, 

 from the more westerly group of Ligiep. The 

 bread-fruit tree is wanting in Ligiep, and the cocoa- 

 tree does not rise above the rest of the wood. 

 Temo, half way to Ligiep, is a small desert island, 

 where those who undertake the voyage pass tlie 

 night. Mesid, a single detached island, lying east- 

 ward, about two miles in its greatest diameter, did 

 not afford us, on the lee side, where we approached 

 it, the prospect of any abundant vegetation. Only 

 single cocoa-trees rise from its centre, and the fresh 

 water which they offered us to drink was extremely 



L 2 



