SUBANU PHONETICS AND COMPOSITION MEMBERS. 69 
consonant in our audition where no consonant exists. To vocalize that 
audition we must employ a consonant and thus we take a mere ghost of 
speech and materialize it. If we, long culture ages higher in develop- 
ment with wit and ability (sometimes put to use) of knowing what we 
are talking about, do this in our common speech, think how more potent 
this must be with the rude savages remote in their mountains. It is 
their nature, as it is the nature of most rude folk in the higher cultures, 
to be strong in their speech, and this is most manifest at the beginnings 
and endings of words. We shall examine in another connection the 
mutation of d, a firm and strong consonant when at the end of the word, 
reducible to the weaker lingual effort in medial r when the word receives 
a formative suffix. ‘This principle of strength at either end of the word 
tends to build up the appulse into a true consonant. 
such examination as at this point we may give to the characteristic 
consonant mutations in Subanu is confined to those few instances in 
which we find two forms in use. ‘There are but few more than a dozen 
cases in which mutation is discoverable within Subanu itself, but these 
will prove valuable as establishing an introduction to the larger mass 
of phonetic material which will become available when we discuss this 
inner speech in its relation to exterior cognate languages and particularly 
to its immediate neighbor the Visayan. 
At present we register a note of a distinctive phenomenon which 
not yet are we prepared to comprehend: every single instance of muta- 
tion which we may establish upon purely Subanu material is found in 
connection with the tongue, with two exceptions. ‘These two are wholly 
anomalous; the former is daromog as a variant of domomog, mutation 
from labial nasal to lingual semivowel; the latter is palad as a variant of 
palag, mutation from surd lingual mute to sonant palatal mute. It 
will be observed that in each of these instances the mutation is extra 
seriem and that there is movement out of class, nasal to semivowel in 
one case, surd to sonant in the other. ‘Those who have accompanied 
me in my studies of the Polynesian phonetic will have no difficulty in 
finding in the mutation of palad to palag an instance, solitaryin Subanu, 
of the kappation of t which is so marked a present character of many 
languages in the central and eastern region of the Pacific. 
The lingual mutations which we have been able to segregate for 
study are most frequently from the mute, the maximum speech effort of 
the tongue, clear across the whole range of its activity to the minimum 
effort in the liquid semivowel. ‘The surd lingual mute t affords one 
instance: in posoloron, from the stem solot, we find the mutation t-r. 
This is the weakening of a consonant strong when final into the liquid 
when it ceases to be final upon the addition of a formative suffix. 
The same principle is active in the case of the sonant lingual mute 
d as a final with mutation to r in the following instances: guicoran and 
poguingcora, from the stem cod; linonsoran, from the stem lonsod; pego- 
