Evidence which points strongly to an Indonesian origin of the aborigines of Formosa exists in certain of their articles of handicraft, notably the peculiar Indonesian form of loom, the nose-flute, and the musical bow. (To these I shall refer at greater length under the head of ARTS AND CRAFTS.) Also the custom of certain tribes--notably the Yami, of Botel Tobago--of building their houses on piles.[50] This in a climate, and under conditions, where there is no material need for such construction. When asked the reason for this, one gets the reply customary to any question that one may be foolish enough to ask as to the “reason why” of any custom whatsoever, viz. “Thus have our fathers done.” To my mind, however, the strongest evidence showing Proto-Malay, rather than Chinese, Melanesian, or other affinity, is supplied by the language--considering the dialects collectively--of the aborigines. [Illustration: MEN OF THE BUNUN TRIBE. _Japanese policemen in background._] [Illustration: YAMI TRIBESPEOPLE OF BOTEL TOBAGO IN FRONT OF “BACHELOR-HOUSE.”] I am aware that the evidence of linguistic affinity as in any way indicating that of race is rather disregarded by many anthropologists, on the ground that contact--commercial or otherwise--between peoples often affects linguistic interchange, or results in the introduction of words from the language of one people into that of another. With this I strongly agree, as regards different races living on the same continent (the different races of Africa being a case in point); or even as regards people living on neighbouring islands. With the Formosan aborigines, however, there has been no contact within historic times between themselves and other branches of the Malay or Indonesian race. They themselves are not a seafaring folk, and the people who have invaded their island--certainly since about the sixth century A.D., when Chinese records first speak of it, during the Sui Dynasty--have been successive waves of the Chinese themselves, the Dutch, the Spanish, possibly the Portuguese, and the Japanese. In spite of this fact, the language to which the Formosan dialects show closest affinity is Malay proper, that spoken on the Malay Peninsula, although there is some resemblance to that spoken in Java, judging from Malayan and Javanese words given in books, such as Wallace’s _Malay Archipelago_. It has been estimated that about one-sixth of the words of the various Formosan dialects, i.e. those spoken by the different tribes, have a direct affinity with the Malayan language--that spoken by the Malays proper. With so large a proportion of words bearing a close resemblance, and taking into account the centuries-long isolation of the Formosan tribes--as regards contact with other Malay or Indonesian peoples--there can be little reasonable doubt that the languages have sprung from a common stock, as probably the races have done.