ON THE LANGUAGES OF OCEANIA. 349 
quite possible that the Malays, who are recent settlers in Indo- 
nesia, found the ancestors of the Polynesian race in possession of 
that Archipelago, and adopted these words from them. This 
would explain how it happens that, while they have these few 
words in common, the two races are in other respects so very un- 
like. My examination of the Samoan dialect has led me to be 
strongly of opinion that the basis of it is Indian, and, as to many 
of its simple root words, Aryan; and so I do not see how a Mon- 
goloid people and speech—the Malay—can have produced the 
Polynesians who are essentially Caucasian. But, if a Mongol race 
long ago came down upon the lands of the Eastern Peninsula and 
ultimately got hold of the adjacent islands, driving from them an 
earlier Caucasian race, I then understand how the Caucasians were 
compelled to flee eastwards into Polynesian, and how a small por- 
tion of their speech has remained among the people that supplanted 
them. In support of this view of the question, I purpose to show 
here that many of these so-called loan-words are not Mongolian 
but Aryan in their origin, and that therefore they probably be- 
longed to the Polynesian speech before they became Malayan. 
This view has, in a general way, been long floating in my mind, 
but it has been strengthened recently by contact with a Dictionary 
of the Pali language. The Pali is the sacred language of the 
Buddhists ; for it is the language of their sacred writings. The 
Pali is one of the Prakrit or common Aryan vernacular tongues 
of India, and is, of course, a sister tongue to the polished Sans- 
krit of the Vedas. It Was a language spoken, in the sixth century 
B.c., and in the very region of India where the prophet Cakya 
Mouni himself taught his disciples the new faith. It has now 
been a dead language for over two thousand years, and this is an 
advantage to us in our present inquiry; for, hke Hebrew, these 
Pali words have been stereotyped in their original form by the 
sacred writings, and can be referred to'as unadulterated evidence. 
And, first, I take up some of the pronouns; for the pronouns. 
and the numerals are considered sure tests of kinship. And for 
convenience I will use (1) to mean Samoan, (2) Malay, and (3) 
