246 , THE MOLUCCAS. [chap. 



Wallace as in use in Ceram. The tree itself furnishes almost all 

 the necessary apparatus. One of the large leaf-stalks ensheathing 

 the trunk at the base, when placed horizontally and supported two 

 or three feet above the ground, forms a deep trough, at one end of 

 which is a sieve or strainer. This is very ingeniously constructed 

 from the dense mat of dark fibrous substance found at the base of 

 the leaf- stalks. Scraped and teased out until of the necessary 

 mesh, it is kept stretched tense by means of an elastic stick bent 

 over and fastened to the top. The crushed pith is mixed with 

 water in the trough, stirred up and down, and pushed toward this 

 sieve. Passing this, it runs along another palm-stem gutter, which 

 is blocked at the farther end by a handful of the same con-like 

 substance as that forming the first strainer, and the liquid then 

 falls into a canoe, slightly tilted so that an overflow of tolerably 

 clear water takes place at one end, while the sago is deposited as a 

 pasty substance at the bottom. The raw starch thus obtained is 

 afterwards dried and baked into small cakes or biscuits five or six 

 inches square, which are tolerably palatable and well adapted for a 

 traveller's use, owing to their portability and the amount of 

 nourishment they contain, but the flavour is very dift'erent from 

 that of European sago. The double washing, granulating, and 

 roasting which the latter undergoes renders it more attractive in 

 appearance, but robs it somewhat of the characteristic taste of the 

 native article. 



We had obtained but two specimens of the Kicobar pigeon 

 during our ^'isit to Obi, and hearing from a native that the Weda 

 Islands, upon which he had once landed while on a fisliing ex- 

 peditiou, abounded in these birds, we determined to \isit them on 

 lea\dng Batchian. They are an uninhabited group of coral islands 

 lying a few miles off the southern extremity of Gilolo, very 

 numerous, but of no grreat size, the largest — as far as is known — 

 being not more than four or five miles in length. Theii- position 

 being purely conjectural, we approached them with caution, but it 

 was to all appearances unnecessary, for there were no signs of 



