110 MADRAS FISHERIES BULLETIN VOL. Jtll, 



insignificance compared with two other facts. Of these the first is 

 that outrigger canoes of essentially the same design but often with 

 double frames instead of single are employed in Madagascar, and 

 on the East Coast of Africa; the second, that in the Marquesas 

 islands a single outrigger of almost identical construction is used 

 commonly for inshore fishing.* As I suggest below, the general 

 inference from the far eastern distribution of the different 

 types of outrigger canoes and their present-day builders, tends 

 to show that at one time the Polynesian users of the single 

 outrigger were distributed throughout Indonesia, that the incursions 

 of Mongoloid people — the main stock from which the present Indo- 

 nesians are derived— displaced the Polynesians and drove them 

 afield both east and west. It may well have happened that the 

 Marquesan type referred to above, wherein the stanchion connexion 

 between the booms and the float is on essentially the same design 

 as that used in the North Java type, was in use by the aboriginal 

 inhabitants of Java on the coming of the proto-Malaysians ; 

 that the dispossessed, some of whose descendants may be the 

 Marquesans of to-clay, carried the design with them in their wan- 

 derings, and that the incomers at the same time adopted the design 

 and transmitted it with little change to the present inshore Java- 

 nese fishermen. It is very probable that in the days when the 

 double outrigger design was dominant in Indonesia for the largest 

 craft as well as for small dugouts, that the larger boats using 

 the North Java design employed it in the double condition; 

 in this stage. I believe, it must have been when one or other 

 of the waves of Malaysian emigration struck the shores of Mada- 

 gascar, and introduced it there. As the Malagasy languages even 

 in the case of the latest arrivals, in regard to the inclusion of 

 Sanskrit words, aie purer and in a more archaic condition than 

 either Malay or Javanese, although otherwise closely related there- 

 to, we are able to date back the time of the use of the present 

 North Java type in its own country to a period anterior to the date 

 of the first Buddhist or Brahman missionary influence in Java, 

 i.e., to a time prior to the Christian era. It is very probable 

 that both the double and the single form of this type were in use 

 concurrently in Java, just as we find both forms in use in 



* Alexander, A. B., " Notes on the Boats, Apparatus, and Fishing Methods employed 

 by the natives of the South Sea Islands." Report of the U.S. Fish Co»i»iissioner for 

 /<?o/, Washington, 1902, p. 745. 



