S2 The Oceanic Languages Semitic : 



participial r)i is changed into 5, v, and wh, and like the My. 

 fn in 7)%a is inseparably attached to the a, that is (as in the 

 Mod. Syr. causative) only the participle is used. 



2. We have the very same as this Fa. ha, Fi. and Mota va> 

 Ma. wha, in An. iirni, Ero. ainp%, Mg. anfipl, amijpa ; but to 

 this also Mg. prefixes the participial m. The same is found 

 sometimes in Papuan, Araga as ma va (Codrington, " The 

 Melanesian Languages," p. 187.) It is really doubling the 

 participial m, though unconsciously. 



3. In Mg. niaha we have the participial m as before 

 separable, a the causative prefix, and the verb substantive 

 liv (see above § 4, a., and § 6, h), so that aha means make to 

 he. In Papuan and Ma.-Ha. this m being as before insepar- 

 able and changed to 6, /, and v, and wli, in haka, faa, vaha, 

 whaka, vaga, and in Mg. itself, as in 2, inseparable and 

 changed into m2X, as in ampa, in ampctha. Again as in 2, 

 before ampa, so before onajxiha, Mg. admits the participial '??!'. 



h. The above may be thus shown : — 



1. The simple cans. Oc. prefix Mg. a. Lifu, Mare a. 



2. This a with the participial m (changed to h, f, v, wh, 

 7np, &c.) as Mg. aonpa, An. ioni, Ero. ampe. My. ma. Fa. ha 

 (hi), Fi., Mota va ma, %vha. 



3. The simple a prefixed inseparably to the verb substan- 

 tive Mg. aha, Rarat. aka. 



4. The a with the inseparable m, in 2, prefixed inseparably 

 to this verb substantive, Mg. ampaha, Fa. haka, faka, Fi. 

 vakaj, Mota vaga, Sam. faa. Ma. whaka. 



Note 1. — Perhaps Rarat. aka belongs to 4, not to 3. 



Note 2. — While the above as compared are etymologically 

 identical, allowance must be made for difference of use. 



Note 3 — In Mg. and My. man, mang, Ja. an, ang, the 

 n, ng, may be roughly described as euphonic, though, as will 

 be seen below, they are not perhaps purely euphonic. The 

 otlier phonetic changes are mainly those of m (participial) to 

 h, f, V, ivh, and mp. The a also appears sometimes, but 

 rarely, as ^ or e. 



c. Comparison : see Semitic Grammars. 



The causative is formed in Syr., Assy., Arb., and Eth., by 

 prefixing a (sometimes weakened to e, I), this has been 

 softened from ha, of which the h is retained in Sab. and 



