VII.] THE OCEANIC MONGOLS. 251 



these words, many of them quite ordinary terms in daily use, 

 could not all have been left behind by the Malayan settlers in 

 Madagascar had the migrations taken place within the last 2000 

 years or so. But none, absolutely none, are found in Malagasy, 

 which language must therefore have crossed the Indian Ocean in 

 pre-Hindu, that is, remote prehistoric times. 



The same inference follows from a critical study of the Arabic 

 elements in Malagasy, which have misled so many 



J \ m The Arabic 



observers, and even given rise to the theory that Elements 



, , , , . r ^i Prehistoric. 



the Madagascar tongue is a corruption ot the 

 Arabic 1 ." A less extravagant, but no less mistaken view, still 

 prevailing in some quarters, assumes that the Arabic words were 

 all introduced either directly through the Muhammadan Arabs, 

 or indirectly through the Muhammadan Malays, from which it 

 would follow that the immigrants from Malaysia were after all 

 the historical Malays arriving since 1000 B.C. (Oppel), or even 

 "probably not over 200 years ago 2 ." But Mr J. T. Last, who, 

 I think rightly, identifies Madagascar with the island of Menuthias 

 described by Arrian in the third century A.D. 3 , suggests the 

 " possibility that Madagascar may have been reached by Arabs 

 before the Christian era." This " possibility " is converted 

 almost into a certainty by the analysis of the Arabo- Malagasy 

 terms made by Dahle, who clearly shows that such terms " are 

 comparatively very few," and also " very ancient," in fact that, as 

 already suggested by Prof. Fleischer of Leipzig, many, perhaps 

 the majority of them, " may be traced back to Himyaritic 

 influence 4 ," that is, not merely to pre- Muhammadan, but to 



1 Dr Vanderkamp quoted by the Rev. L. Dahle, Antananarivo Annual, 

 1876, p. 75. 



2 Col. Maude, Jour. Anthrop. Inst. 1896, p. 71. 



3 "His remarks would scarcely apply to any other island off the East 

 African coast, his description of the rivers, crocodiles, land-tortoises, canoes, 

 sea-turtles, and wicker-work weirs for catching fish, apply exactly to Mada- 

 gascar of the present day, but to none of the other islands " (Jour. Anthrop. 

 Inst. 1896, p. 47). 



4 Loc. tit. p. 77. Thus, to take the days of the week, we have: Malagasy 

 alahady, alatsinainy ; old Arab. (Himyar.) al-ahadu, al-itsndni ; modern 

 Arab, el-dhad, el-etnen (Sunday, Monday), where the Mai. forms are obviously 

 derived not from the present, but from the ancient Arabic. From all this it 

 seems reasonable to infer that the early Semitic influences in Madagascar may 



