times before that could 'happen', if he did not 'make it happen' in the manner intimated above. No doubt he generally began working on the haphazard principle; but after having destroyed his sikìdy several times and begun anew—just sufficient to make his spectators understand that it was a very serious affair—he had resort to artificial means, and made it succeed. I fancy this was the general practice in producing the charms described above."
Mr. Dahle thinks that the practice of sikìdy among the coast tribes is not so fully developed as that in use in the interior of Madagascar, except, possibly, in the district of Màtitànana (S.E. coast), for here there was an ancient Arab colony, and a great many Arabic customs have been retained by the Antaimòro, as well as by the Antanòsy, further south towards Fort Dauphin, where Flacourt was governor.
The Bétsimisàraka have, besides the systematic kind of sikìdy already described (sikìdy alànana), at least six other kinds. These are said to be much simpler than the ordinary kind of divination; one, for instance, has only two columns or row; another kind, also with two columns, is worked by using in some cases three beans, as well as one or two. Other kinds, although styled Sikìdy kofàfa or vèro, can hardly be properly called sikìdy at all. The procedure is simply the following: You take an indefinite number of kofàfa or vèro (kofàfa, a broom made of grass stalks, vero, a tall grass), and you then take out two and two until you have only one or two left. But you must have settled in your own mind at the outset whether one left shall mean good luck, and two bad luck, or vice versa. A similar practice is, we know, found among Europeans also, but only as an amusement.
There is, says Mr. Dahle, another kind of sikìdy (if we like to call it so), which, I have been told, is practised by an old woman in Antanànarìvo. Something had been stolen and nobody knew the thief, but they suspected he was to be found among the servants. So the old
