wholly or partially, round those who remained behind, and to recede farther and farther from the centre under the pressure of later emigrants, until they at length arrived at the edges of the earth. And here, even at the present day, their original ring might still be found, but broken through in many places by Aryans, Chinese, and other more central nations, like as the external terrestrial strata are broken through by the internal Plutonic rocks. Among these more central nations would probably be included those Americans who employed the s five, and who are now usually called the Red Men. They would enter America, like the earlier settlers, at the north-west, but only to pass through so desolate and unattractive a region without staying to form settlements. Leaving on their left the Arctic tracts, and on their right the Rocky Mountains, they would eventually reach the great lakes, and then spread themselves abroad in the basin of the Mississippi and over the Alleghany mountains, until they gained in the end the shores of the Atlantic. Oregon, California, and Central America may have been invaded later by them from the Prairies.

To return to the external races and their numeration. The k-m hand, which seems employed to form the prevalent 'ten' of Ethiopian Africa (as Africa south of the Great Desert may be styled, in contradistinction to Northern or Libyan Africa), would probably be composed, as in other cases, of two words for 'finger'. At least it would be so, if we may form a judgment from the following Hottentot words: :

=

t'koam, 'two' 'finger' (t'koa or t'ko) + 'finger' (m or am). t'koam, 'hand', 'fingers',

=

='finger-finger'.

The process would be precisely the same as in the Sioux words cited above (p. 6), where 'hand', napai, = 'fingers', naap, = 'two', nopa. The first of these two African 'fingers'

may be reserved for future notice. The second, m or am, is to be recognised in several African languages, such as those which follow, where the various suffixes may be mostly traced to words for 'arm' or 'leg':

These Ethiopian or Trans-Saharic words for 'finger' and 'toe' might be employed also to explain the Basque amar or ama-, 'ten', at least if there were any reason for inferring the existence of an Ethiopian element in the Basque language; and, should such an element be discovered, there would be the less reason for supposing the Finnish 'tens', kümme, kymmen, and kämen (p. 24), to bear only an accidental resemblance to the Ethiopian 'tens', kumi, guma, and gomen, and to the kindred Ethiopian words, guma, 'bracelet', and also 'heel', gumen, inner hand', and komen, 'foot-sole'. Here, then, a short digression from the subject of numerals may be allowable, while we consider, with especial reference to ancient Spain and Ethiopian Africa, a few of the other terms in which primeval affinity, if it exists, may be expected to be betrayed.

In ancient Spain there co-existed three races or nations, all of which may have left their mark on the Basque language. There were the Celts, probably the latest settlers of the three in point of time, who were widely and perhaps thinly scattered over the northern, western, and central regions, and who may be presumed to have entered the country from Gaul. In the second place, there were the

Iberians, who were purest along the Mediterranean and the slope of the Pyrenees, and may have come from the east, originally from the Caucasian Iberia. Finally, there were the Cynetæ, in the west or south-west, who might possibly either have come from Africa, as their position would suggest, or else have formed part of an Ethiopian ring once encircling the Old World. In either case, should the Cynetæ prove to be Ethiopians, it would most likely be Libyan or Sub-Semitic intrusion from the east into Barbary and the Sahara that divided them from their kindred beyond the Desert.

The following resemblances between Trans-Saharic languages and Basque are in favour of the hypothesis that the Cynetæ were Ethiopians, or, at least, that there was an Ethiopian race in ancient Spain :—

The Central Indian (Kol, Santali, Bhumij, Mandala) buru, 'mountain', should likewise be compared with the Basque buru, 'head', as also the African (Kiriman) muru, 'head', should be with the Basque muru, 'hill'.

The following words for 'hair' are found in Madagascar and Australasia, regions where the Ethiopian or t-n hand

five has been previously detected (p. 21):-Madagascar, wulu; Lombok, bulu; Sumbawa and Celebes, welua; Borneo, ulu; Australia, yal, eeal, (and also walo, wollar, wollak,

Sudo is nose' in a Finnish language, the Mordvin, where kämen isten', and thus resembles Ethiopian 'tens' (p. 24).

In Central India, á, bái, váyi are words for mouth'; and in Southern India, vai, bai, boi: in the Caucasus there is the Tshetsh bagga.

To these should be added the following words for 'tongue':-Tasmania, mena; Polynesia, mangee (Paumotu),

mea (Vanikoro), mia (Tanema).

With regard to the value of these words as signs of affinity in language or in race, it may be mentioned that all their English exponents are German as well, thus indicating our true pedigree; and that they form, moreover, ten out of the twelve names for members of the human body that have been selected as tests of ethnical affinity (Latham, Comp. Phil., p. 679). The other two names are, in Basque-escu, hand', for which see ante, p. 16; and azur or ezur, 'bone', which comes near the Munipuri surru, 'bone' (between Assam and Burmah), the Dungmáli súr-wá, 'bone' (E. Nepal), and the Gyarung syárhú, 'bone' (Tibet).

Of the two great divisions of the globe, ‘earth' and 'sea', the first is called in Basque lur, which seems the same as the Gaelic lar, the Welsh llawr, the Cornish ler, and the Breton leur, which have all the like meaning; while 'sea' is called itsaso, which may be explained without difficulty from Guinea dialects :

ENGLISH.

Salt.

Water.

--

Sea.

The double meaning of the Greek aλs, salt, sea', would lead us to identify together the Avekvom etsa, salt', and etya, sea'. Similar African words for 'salt' are:-yase (Koama), yesa (Guresha), adsi (Papel), and several more. It is plain that such a word as the Basque its-aso, ‘sea', might be made out of the Avekvom etsa-esonh, the Adampe edse-esi, the Mahi idse-ezi, or the Dahomey dse-zi, saltwater'. 'Water', again, is su in Turkish; wesi in Finnish ;