230 



DISTRIBUTION OF PLANTS. 



7. Kawaki coral-island. In plain sight from Raraka, at the dis- 

 tance of *' ten or twelve miles," another narrow fence-like rim of land 

 commenced encircling a wide expanse of ocean ; this proved to be 

 the coral-island of Kawaki, measuring " sixteen miles by ten." 



On the 2d of September, Mr. Brackenridge landed with others on 

 Kawaki ; but found there " only plants already seen by us on other 

 coral-islands," as the 



(Procris ?) or "Morus?" (compare No. 1 Raraka coral-island, &c.).* 

 Gea. " Rubiac. berried," (compare Taiara coral-island). 

 (Pemphis) or Glen. " Lythr." (compare No. 1 Taiara coral-island). 



8. Aratika, or Carlsholf coral-island. Sailing thence about " twenty 

 miles," the Vincennes towards evening arrived in sight of Aratika ; 

 having its Western extreme in " S. Lat. 15° 26' and W. Long. 145° 

 40'." This coral-island measured "eight miles long by five wide;" and 

 at one time, was nearly all included within the range of vision from 

 the mast-head. In sailing along, the rim of the island was observed 

 to be chiefly a wash, with a succession of dry banks, none of them 

 sufficiently elevated to be crowned with vegetable growth. North of 

 this low portion, a clump or dense grove of trees commences, and 

 extends uninterruptedly more than a mile, or as far as the ship pro- 

 ceeded. 



On the 3d, Mr. Brackenridge and myself landed in front of the 

 Watering-place; and entering the woods a fourth of a mile, were sur- 

 prised to find a large and deep pool of fresh-water, fifty or sixty 

 feet in diameter. The grove was more forest-like than any we 

 had previously seen on coral-islands ; the ground being covered 

 throughout with a layer of dead leaves. The principal tree was the 

 Calpidia, forty feet high, with the trunk three feet in diameter, often 

 crooked and irregular, branching nearly from the base, and presenting 

 besides some anomalies of growth ; the branches being sometimes com- 

 pressed, after the manner of the exposed portions of the roots, or 

 turning downwards and entering the ground, giving rise to upright 

 secondary trees. The Psilotum was abundantly attached to the 

 trunks ; and the Polypodium rambling over them in luxuriance, as 



* ('ocos nucifera, (bis Clermont-Tonnerre to Raraka coral-island). Two or three cocoa- 

 palms were the only signs of inhabitants met with, even by the party that landed : the 

 " people" known to the natives of Raraka, living doubtless on a distant part of the 

 island. 



