92 THE SUBANU. 
These are confirmatory of the deductions which have been drawn 
from the Subanu material; in the first and third of these inverts there 
can be no doubt whatever that the initial syllable is inverted as a unit, 
in the second invert we shall find justification in regarding the inversion 
as of the same type. 3 
This theory of inversion as dealing with syllable units comports 
with our broader comprehension of languages so primitive as are these 
of agglutination. While the word-stem is frequently dissyllabic we 
must regard it as compact of monosyllabic roots. In the languages of 
isolation the two roots which enter the stem are capable of independent 
existence and most commonly are found free in the same speech. In 
agglutination the secondary root has in most cases ceased its free exist- 
ence and in the process of such disuse has undergone more or less of 
form-change, so that it has become merely a composition member. 
The principal root remains susceptible of necessary modification as a 
speech unit. I may note the occurrence of this type of inversion, 
though infrequent, in the isolating languages of Melanesia. 
It was not within my original plans for the scope of this work that 
the collation of the Subanu affiliation should extend beyond the im- 
mediately circumjacent Visayan. It was easy to recognize that in 
the many languages of the Philippines many interesting discoveries 
might be made and that more extended study must be fruitful in valu- 
able results. After due consideration I determined to relinquish this 
study to those whose concern is more specifically directed to Philippine 
linguistics and to those masters of Malay philology who may be ex- 
pected to deal with the new material which has been given me to arrange 
and to order for their examination. It will be understood that my 
particular object has been to sift this Subanu for such data as might be 
found to bear upon my own specific study of the early phases of the 
Polynesian speech. With that I am quite content. 
But it chanced that while these pages were being put into type 
my attention was somewhat fortuitously directed to Dr. Seidenadel’s 
study of the language of the Bontoc Igorot. Immediately I recognized 
a marked similarity in parts of the vocabularies of the tworaces. ‘They 
are widely separated; almost the whole length of the archipelago lies 
between the Igorot of Bontoc in the northern tip of Luzon and the 
Subanu of the southern extremity of Mindanao. Despite the dis- 
tance which parts them they have one condition in common: each 
is interiorly situated with reference to a Malayan people of more 
advanced culture and richer development; the Subanu an inclusion 
within the Visayan area, the Bontoc Igorot within the Tagalog region 
of predominance. 
Despite linguistic differences, this condition is readily compre- 
hensible in our acquaintance with the Philippines. The Aetas and others 
of the true negritos represent the survivors of a primitive autochtho- 
