SUBANU-VISAYAN FILIATION. 91 
So far as relates to the former group, we might conclude that the 
process was an interchange of the initial consonants of the former and 
the latter syllables, regarding the initial syllable in each case as open, 
that is, consisting of consonant and vowel. In the second group we 
should have, in continuation of the theory that the interchange is 
between one syllable and another, to regard each syllable as closed and 
that the movement applied to final consonants. This is by no means 
satisfactory; it involves a complication which is foreign to language of 
so elemental a type. 
Our next group of three will afford us a better view of the process. 
sora 
isda 
luma 
odma 
gonlo 
onglo 
Subanu 
Visayan 
Here we have no difficulty in seeing that the syllabification is 
gon-lo, lu-ma, so-ra. ‘This makes clear what has happened, the initial 
syllable has been inverted without any doubt in Juma and sora, lu-ma-ul- 
ma-od-ma, so-ra—os-ra—ts-da. In the first group of three we find the 
same syllable inversion in all three cases; gos-od—sog-od-sog-ot. In the 
second group it holds; t-mod-—ti-dom-ti-gom. 
There remain now three apparent anomalies, gonlo, tondong, tung- 
dong. Instead of nog-lo we find ong-lo; instead of ton-ngod ton-god; 
instead of tung-ngod ton-god. Since all involve the palatal nasal we 
may safely conclude that the same principle of inversion of syllables is 
operative, but that the result is subjected to perturbation produced by 
some attractive quality in the palatal nasal, a closure which we have 
established as among the first of the consonant possibilities to be 
developed. 
Our material affords us a group of four vocables, in which at first 
sight there appears to exist a different type of inversion, the interchange 
of consonants concurrently brought together. ‘They are these: 
tondo 
todlo 
sopla 
sompa 
ondoc 
hadloc 
ondao 
adlao 
Subanu 
Visayan 
As these have arisen for consideration in the foregoing text I have 
indicated a more satisfactory explanation, that when for any reason a 
liquid is assumed by the stem the preface of the series nasal is dropped 
by the stem mute; or, that when a stem liquid is dropped the mute is 
prefaced by way of compensation. Just which of these two statements 
is the proper view must await the determination of the true stem in 
these words and that can be accomplished only by following them 
through their various occurrences in Indonesia. 
The Bontoc Igorot affords us three instances of inversion which 
may properly be adduced for comparison. 
niug 
inyog 
lipay 
paley 
Subanu 
Bontoc 
gosa 
ogsa 
