I have used, compared and brought together all these loose materials, and thus succeeded in restoring about 234 words of this language, a list ample enough for all historical purposes. This contains besides 50 words of the Eyeri and eastern dialects, with 38 of the Cuban or western dialects, useful to show the variations of dialects. We know that from Bahama to Cuba, Boriquen to Jamaica, a same language was spoken in various slight dialects, but understood by all: Columbus himself says so.
But this language, which had also partly spread in Florida, and in South America, has the appearance of being a mixt speech. This appears from the many synonyms, the deviations of dialects, and the double forms, or relative position of words. In the small eastern islands the _Eyeris_ or _Cabres_ had been destroyed by the Caribs, who preserved the women, and these preserved their own language, mixt with some Carib words and taught it to their daughters; so as to produce a double language, that of the women being quite peculiar. This singular fact well authenticated, will enable us to presume a similar conquest and custom, wherever we shall meet in America, with a peculiar female idiom.
The many nations or tribes mentioned in the traditions, which had gradually amalgamated; and the settlement in Cuba and Hayti of the Mayas, will account for this mixture of synonyms. But the existence also of many homonyms, leads us to a former more simple speech, probably monosyllabic and quite regular as the oriental idioms, to which it is most akin.
From the primitive languages of North Africa and South Europe, it had received this regular position of ideas; but by the mixture with the Maya and Mexican nations using the inverse form, it borrowed that new form. The same happened in Europe to the Celtic and Oriental tribes, who received in Greece and Italy the inverse form of speech from the Scythian, Illyrian and Gothic tribes.
The comparative examination of the Haytian and dialects, was pursued by me, _upon all the languages of the earth_, as I was determined that one American nation at least, should be traced philologically to its real origin. Thus I found many thousand analogies of it, out of which I have used about 1500 in the annals, notes and vocabularies. A single American language does then contain more comparative analogies in about 200 words than all those collected by Vater and Malte, out of 400 American languages; and this fact upsets all the illusions, theories and false views, based thereon by them, Humboldt and others.