THE ADMIRALTY ISLANDS. 467 



some of the crew often sit. The canoes are from 30 to 40 feet 

 in length. 



The sail is nearly square in form. It is hoisted to the top of 

 the mast, and set so that one corner is uppermost. The opposite 

 corner does not nearly reach down to the canoe, hence the square 

 sail being high above the water has a very peculiar look when 

 seen over the sea at a distance. As at all Pacific islands, 

 apparently the outrigger platform is the place of honour, and 

 the seat of the head-man or chief. Oto, the chief of Wild 

 Island, never occupied any other position, and never touched a 

 paddle. 



Small canoes with single outriggers, holding one or two per- 

 sons, are used for paddling about the reefs round the islands. 

 The large canoes are manned by from 10 to 15 men. 



The natives swim hand over hand. They never take a 

 header in diving, but jump in after anything upright, sinking 

 feet first with the body inclined forwards. 



Long sein-like nets are used for fishing. These nets are 

 probably the property of a community, for they are kept hung 

 up in the temples. I saw one about a fathom in depth and of 

 very considerable length. Hand nets fixed on elbow-shaped 

 frames of wood are also used. Stake nets are used, and lines of 

 stakes are conspicuous objects just off the shore near the 

 villages. 



Fish-hooks are used made of Trochus shell, all in one piece. 

 They are of a simple hooked form without barb. The natives 

 did not seem to care for steel fish-hooks, and apparently did not, 

 at first at least, understand their use. It is possible that they 

 have never found out the plan of using bait on a hook. All 

 Polynesian and Melanesian fish-hooks which I have seen are of 

 the nature of artificial baits of bright nacre, imitating small fry 

 in the water. If the natives did not understand the use of baits, 

 it is no wonder that they despised European fish-hooks. 



The tool in most constant use by the natives is a small adze, 

 consisting of a natural crook of wood with a Terebra maculata shell 

 bound on to it, the shell being ground down until only one lateral 

 half of it remains. Such small shell adzes were abundant enough 

 still, but in most cases the shell had been replaced on the handle by 



hh2 



