Again—the Negros themselves are referable to an extreme rather than a normal type; and so far are they from being co-extensive with the Africans, that it is almost exclusively along the valleys of rivers that they are to be found. There are none in the extra-tropical parts of Northern, none in the corresponding parts of Southern Africa; and but few on the table-lands of even the two sides of the equator. Their areas, indeed, are scanty and small; one lies on the Upper Nile, one on the Lower Gambia and Senegal, one on the Lower Niger, and the last along the western coast, where the smaller rivers that originate in the Kong Mountains form hot and moist alluvial tracts.

From whatever other Africans the Negros are to be separated, they are not to be disconnected from the Kaffres, the chief points of contact and transition being the parts about the Gaboon.

Neither are the Kaffres to be too trenchantly cut off from the remarkable families of the Sahara, the range of Atlas, and the coasts of the Mediterranean—families which it is convenient to take next in order; not because this is the sequence which most closely suits either their geography or their ethnology, but because the criticism which has lately been applied to them best helps us in the criticism of the present affiliations.

On the confines of Egypt, in the oasis of Siwah, we find the most eastern members of the great Berber, Amazirgh, or Kabyle family; and we find them as far west as the Canary Isles, of which they were the occupants as long as a native population occupied them at all. Members of the same stock were the ancient subjects of Jugurtha, Syphax, and Masinissa. Mr. Francis Newman, who has paid more attention to the speech of the Berber tribes than any Englishman (perhaps than any European), has shown that it deserves the new and convenient name of Sub-Semitic—a term to be enlarged on.

Let us take a language in its first state of inflection, when passing from the monosyllabic form of the Chinese and its allied tongues, it just begins to incorporate with its hitherto unmodified nouns and verbs, certain prepositions denoting relation, certain adverbs denoting time, and certain pronouns of person or possession; by means of all which it gets equivalents to the cases, tenses and persons of the more advanced forms of speech.

This is the germ of Conjugation and Declension; of the Accidents of Grammar. Let us, however, go farther. Over and above the simple juxtaposition and incipient incorporation of these previously separable and independent particles, let there be certain internal ones; those, for instance, which convert the English Present Tenses fall and speak into the Preterites fell and spoke—or something of the same sort.

Farther still. Let such changes of accent as occur when we form an adjective like tyránnical, from a substantive like týrant, be superadded.

The union of such processes as these will undoubtedly stamp a remarkable character upon the language in which they appear.

But what if they go farther? or what, if without actually going farther, the tongues which they characterize find expositors who delight in giving them prominence, and also exaggerate their import? This is no hypothetical case.

A large proportion of roots almost necessarily contain three consonants: e. g. bread, stone, &c., pronounced bred, stôn, &c. This is one fact.