102 THE SUBANU. 
material we are entitled to estimate the value of that material. I can 
do nothing better than to quote from Edward Tregear upon this very 
point (Maori-Polynesian Comparative Dictionary, page XIII): 
In this very voluminous work Humboldt examines the vocabularies and 
grammatical construction of the Oceanic languages and considers that the 
Tagal of the Philippines is the leading dialect. His vocabularies, however, 
were of a very imperfect character, and his deductions would have been con- 
siderably modified had he possessed the information at present at our service, 
his Maori being the Maori of Lee and Kendall (1820) and his Tongan, if 
possible, still more defective and illusory. 
I have been at pains to discover what linguistic information as to 
the languages of Polynesia was available to Humboldt and therefore 
through him gave Bopp the data for the creation of this family. A 
very few word-lists were buried in the narratives of the great explorers; 
even if we assume that Humboldt had access to them all the material 
was in very imperfect condition and by nomeanstrustworthy. ‘Tregear 
has characterized the “New Zealand Grammar and Dictionary”’ of 
Lee and Kendall. The Tongan vocabulary was that of Mariner’s 
“Tonga Islands,’’ published in 1818 and filled with errors which at the 
present day are impossible of resolution. ‘The work of Davies on the 
language of Tahiti had been published in 1823, but its present value is 
that of a curiosity. In 1837 Chamisso had published his brief and 
inaccurate vocabulary of the Hawaiian, but it does not appear that it 
affected Humboldt’s work. In these few items we have the sum of the 
data, both scanty and untrustworthy, on which rests the Malayo- 
Polynesian family of speech. 
I arraign this family (experience has proved it a deadening collo- 
cation), upon the following grounds: 
1. That the evidence upon which it is sought to support it is 
incompetent, immaterial, and irrelevant. 
2. That a family of languages can not be constituted of members 
belonging to radically distinct orders of speech, and that in this case 
the Malayan is an agglutinative speech and the Polynesian isolating. 
3. That the use of infixes, characteristic of all the Malayan lan- 
guages and necessary to their use in speech, is wholly unknown to any 
of the Polynesians. 
4. That the Polynesian is essentially a language of open type in 
its present stage and that a consistent effort has been operative to 
excise final consonants in stems where inferentially they existed in a 
remote past; that the Malayan languages admit closed syllables and 
that in very many instances there has been an assumption of conso- 
nants in order to close syllables originally open. 
5. That the fixed element of the Polynesian lies in its vowel 
structure; that the vowels of the Malayan are most uncertain, and that 
the permanent elements are in the consonant skeleton. 
