INDONESIAN ANNOTATIONS ON THE VOCABULARY. 119 
56. man bird. 
33. Manu Bahasa, Ceram Alfuros, | 43. manuti Wayapo. 
Savu, Kisa, Menado,San- | 44. mano Saparua, Lariko, Liang, 
guir, Sula, Morella, Cai- Batumerah. 
marian, Baju, Salibabo, | 45. manok Kei, Banda, Bentenan, Ka- 
Togean, Bouton. yan, Magindano, Subanu, 
34. manuk Malay, Ponosakan, Sulu. Visayan, Bontoc Igorot, 
35. manul Bahasa. Matu, Gah, Matabello, 
36. manue Amblaw, Awaiya. Teor. 
37. manui Kayeli. 46. manoko  Bolanghitam. 
38. manuo Teluti. 47. mani Waigiou Alfuro. 
39. manuol Bahasa. 48. manik Gani. 
40. manuwa_ Ceram Alfuro. 49. monok Dyak, Bontoc Igorot. 
41. manuwan Ahtiago. 50. malu Ceram Alfuro. 
42. manut Massaratty. 51. malok Wahai. 
In Indonesia we find a considerable occurrence of the final k which 
has been twice noted in Melanesia; the forms in which it occurs are 
34, 45-46, 48-49, 51, representing a score of languages. I have sug- 
gested (Subanu, 132), but on less complete material, that the final 
palatalisa part of thestem manuk. It nowseems preferable to regard 
the k as an assumption of a word-determining element under the genius 
of Indonesian speech and not a pertinence of the stem. Thus only is it 
possible to correlate the forms of the type in 1 in manul (35) and manuol 
(39), of the type in t in manut (42) and manuti (43), in n in manuwan 
(41). Of these variants mutation of consonants could be employed 
to explain none save the t forms, and even that is only remotely pos- 
sible. Of the inner consonant we have but a single variant, malu (50) 
and malok (51). ‘The n—-1 mutation is well established in these language 
provinces, and in Melanesia we find one of these forms in 9 malu 
credited to Alite. The former vowel undergoes but one mutation, 
a—o, found in 49 monok, which occurs in the Dyak and Bontoc Igorot, 
both languages of a type but slightly advanced, and the same muta- 
tion has been observed in 26 mon and 27 monmon from Melanesia in 
two particular regions of the northern Bismarck Archipelago and the 
northern New Hebrides. The final vowel appears to be under a 
notable tendency toward mutation, a tendency obscurely comprehen- 
sible as in some way of a compensatory character, a movement which 
is also apparent in the frequent Indonesian assumption of one of several 
final consonants. Of these mutations we find two distinct types. 
That exhibited in the formula u-o is a recession toward the central 
vowel; that exhibited in the formula u-e is a movement to a position 
of similar advance on the other limb of the vowel scale. With these 
clear mutations I include the several instances in which merely ocular 
inspection seems to present cases of an unmodified u plus the assump- 
tion of a new final vowel, as shown inter alia in the case of 38 manuo. 
It appears to me that in such cases the apparent mutation u—uo is really 
the mutation u-o, but that in the course of the change sufficient of 
the u persists to give the impression of a semivowel lightly impressed 
upon the resultant vowel and that the mutation is better expressed 
