volumes, yet they would, one and all of them, be less conclusive on the point, than the fundamental correspondence, both in the words and in the structure of their languages. With but little difficulty, and in some of the instances with none at all, Tahitians, Marquesans, Samoans, Tongans, New Zealanders, and Hawaiians, to say nothing of the less known groups, can render themselves intelligible to each other; and of this similarity of dialects, the strongest as well as the most gratifying proof is to be found in the fact, that native converts of one Archipelago have sometimes gone forth, as missionaries, to communicate the glad tidings of salvation to another. Thus a chief who accompanied Mr Ellis from Tahiti to the Sandwich Islands, often addressed the natives with effect; and Sir Edward Belcher found a little colony of Samoan teachers labouring, or rather wishing to labour, among the savages of the New Hebrides. To offer more specific evidence of the fundamental correspondence in question, the subjoined table is quite decisive, at least with respect to the words; for the identical meanings of six nearly identical sounds, in three different dialects, cannot possibly be accidental.

Perhaps a careful examination of different dialects might suggest some hints as to their comparative antiquity. As the general tendency of language to become softer by change would derive special force from the feeble and almost childish organs of the race under consideration, any dialect might reasonably be deemed more recent, in proportion as its alphabet and pronunciation might be more meagre and effeminate. Now the common language of the Polynesian Isles appears to have travelled from the west towards the east. Thus the Hawaiians, and apparently the Tahitians also, abhor a concourse of consonants, while the New Hebrides have their Erromanga, the Feejees their Banga, and the Friendly Islands their Tonga, or, to use the better known name, their Tongataboo. If an Hawaiian were desired to pronounce any one of these three words, he would either insert a vowel between the two consonants, or omit the harsher consonant; and he would most probably adopt the latter course, just as he would transform England into Enelani. In all probability, Tonga and Tona or Kona, the name of a district, already mentioned, of Hawaii, are one and the same word; and, to give an instance of which there can be no doubt, tangata, the Samoan word for man, has been softened into the Hawaiian tanata or kanaka. Again, the very name of Samoa, the chief of the Navigator Isles, involves the letters, which the Hawaiians, as also I believe the Tahitians, alto

gether reject, as being too much for their powers of utterance. Thus they change fashions into patena or pakena; missionary into mitinary or mikinary; and consul into tonatele or konakele. Finally, the Marquesan or Tahitian dialects, though they partake, in an eminent degree, of the softness of the Hawaiian, have yet retained at least one consonant, namely f, which it has discarded. The Fatuiva, one of the Marquesan Isles and Paofai, a chief of Tahiti, would, in the mouth of an Hawaiian, respectively become Patuiwa and Paopai; while there can be no mistake as to the original orthography, inasmuch as the ƒ is distinguished in the one word from v, and in the other from p. Might not a similar application be made of the table which preceded this paragraph, with respect to these three dialects? In the first four of its six words, the v of the Tahitians and Marquesans becomes the w of the Hawaiians ;—the former being, of course, a consonant; but the latter, however it may be classed by grammarians, being really oo, sounded as quickly as possible. If there be any truth in these desultory and incomplete suggestions, then must this Archipelago have been peopled after the Marquesas and the Society Islands, and they again after the more westerly group.

This result, which, so far as the Sandwich Islands is concerned, agrees with the traditionary lore of the Archipelago, is consistent with nearly all the arguments which can be brought to bear on the subject. Looking on the map at the tolerably continuous chain of islands and groups of islands, from Sumatra to the Marquesas, and at the comparatively open ocean between this its last link and the American continent, a plain man would instinctively infer, at least in the absence of evidence to the contrary, that Polynesia, as certainly as Australasia itself, must have been peopled, not from the new world, but from the old; and he would find his inference materially confirmed by the fact, that, on any and every hypothesis, the isles of the Pacific could have been colonized from the westward long before the eastern shores of that ocean contained a single family of human beings; while, on further investigation, he would confessedly discover vastly more numerous traces of Asia than of America, in the ethnographic characteristics of the Polynesian Isles.

The single, absolutely the only, answer to all this, is the physical fact, that the trade-wind blows from the east along the whole breadth of the route which has been chalked out for the primeval colonists of the islets of this greatest of all seas. Now, in the face of so much direct proof of an Asiatic origin, the evidence in question of an American origin amounts to nothing, unless the difficulty of advancing from west to east in spite of the trade-wind actually amounts to an impossibility.

But, so far from amounting to an impossibility, the difficulty itself, strictly so called, can hardly be said to have existed. As the trades, even at their steadiest, take to themselves a few points of

elbow room, having ranged, for instance, in our own case, as already mentioned, between NE. and ESE., the Polynesian groups, occupying about fifty degrees of latitude, might all be intersected, without any formidable interventions of a foul wind, by one and the same track, starting from the westward; and, even independently of this constant oscillation of the ordinary current of the air, the same result could be still more easily and more directly attained with the aid of the opposite monsoons, which blow, with greater or less regularity, during two or three successive months of the year. Moreover, on such a point, one fact is more conclusive than a score of arguments; and, unfortunately for the partisans of the east wind, all the facts are stubborn supporters of the other side of the question. The inhabitants of each group, in whatever direction their ancestors reached it, think nothing of sailing from its westerly to its easterly islands; and Captain Beechy fell in with several men and women, who had drifted six hundred miles with a large canoe in the very teeth of the general direction of the prevailing trades. But, even if the alleged difficulty amounted to an actual impossibility, the claims of Asia to be the cradle of the Polynesians, though they might be weakened, would yet not be disproved. The westerly gales, which generally blow on either side of the region of the trades, might carry vessels far enough to the eastward, to make the tropical breeze a fair wind to the westward, more particularly if they had started from the more northerly coasts of Asia; and, in fact, one Japanese junk, in December 1832, was driven to Woahoo, with four men alive out of her crew of nine; while, again, in 1839, another was found drifting about half way on the same involuntary voyage, with several individuals on board, the same whom we afterwards saw at Ochotsk, which they had reached immediately from Kamschatka, on their homeward route from the Sandwich Islands.

Farther, if the trade-winds had really rendered a voyage from west to east impracticable, Polynesia would, in all probability, never have been peopled. There is not the least evidence for believing,there is not the slightest reason for supposing,—that the Aborigines of America ever possessed a canoe or any other vessel stout enough to survive the dangers of the intervening ocean, during a voyage which could not, under the most favourable circumstances, occupy less than three or four weeks. All the obstacles of the trade-wind notwithstanding, I should more readily conclude, that the Marquesas colonized Southern America, than that Southern America colonized the Marquesas, so far, at least, as the mere question of navigation might be concerned.

From what country, then, of Asia, did the Polynesians spring? Almost, to a moral certainty, from some point, or rather points, between the southern extremity of Malacca, and the northern limits of Japan, an answer which appears to be corroborated by that most conclusive of all features of resemblance, the similarity of language.

Premising that, in such a case, nothing like identity is necessarily to be expected, for, according to general experience, the human race was diffused over the globe rather by the migration of whole tribes, than by the emigration of parts of them, there seems to be no ground for doubting, that the dialects of Polynesia are connected with the languages of the adjacent coasts of Asia. To say nothing of the admitted fact, that the Chinese residents of the Sandwich Islands pick up the Hawaiian with great facility in a short time, the Malayan tongue is universally allowed to be a striking analogy to the language of the groups of the Pacific. To the eye indeed, and perhaps also to the ear, there is said to be a staggering difference in the predominance of vowels on the part of the latter, and of consonants on the part of the former. This difference, however, is susceptible of a satisfactory explanation. The concourse of consonants in the Malayan arises, in a great measure, from an admixture of the Arabic, which, to a moral certainty, must have taken place long after Polynesia began to be peopled; and, even if the admixture in question had been anterior to the colonization of any of the islands, the concourse of consonants just mentioned would, to a considerable extent, have been nominal, inasmuch as the short vowels of the Arabic are rounded without being written. But, further, the peculiarity under consideration of the language of Malacca, supposing it to have been both original and real, would tend rather to support than to impugn the foregoing views. The Hawaiian has been shewn to embody fewer consonants than the Marquesan or the Tahitian, and the Tahitian and the Marquesan again to embody fewer than the Samoan, to the Fejeean, or the Tongan, or the dialect of the New Hebrides, the taboo of the eastern groups, to add another instance to the instances already cited, assuming the form of tamboo to the westward. Now, on the very same principle, one ought not to be surprised to find that the consonants become more numerous and more harsh as one approaches to the native seats of a language so widely diffused.

To conclude this head with one remark more, if any ethnographic similitudes do exist between America and Polynesia, they may be safely considered as common results of one and the same cause. Though the new world must have received inhabitants from the old across the strait which separates them, just as certainly as if the two were connected by an isthmus, yet it might, in all probability, have received others, and those too, in more regular and continuous streams, along the chain of stepping-stones, which extend from China to the north-west coast, comprehending Japan, the Kurile Islands, and the Aleutian Archipelagoes; and, to shew that this supposition is far within the limits both of possibility and of probability, a Japanese junk, such as has been used since the first settlement of this country, lately found its way to the western shores of the new continent, with a living crew on board, and without the aid of any inVOL. XLIII. NO. LXXXV.—JULY 1847.

D

termediate place of refreshment or of rest. In a word, America and Polynesia appear to have been chiefly, if no solely, colonized from one and the same general region of Eastern Asia.-Narrative of a Journey round the World, by Sir George Simpson, vol. ii. p. 2.

On the Age of the Volcanoes of Auvergne as determined by the Remains of successive Groups of Land Quadrupeds. By C. LYELL, Esq., F.R.S., &c.*

The region of extinct volcanoes of Auvergne derives its peculiar interest from the circumstance of its never having been submerged beneath the sea during a period in which its geological and geographical structure, and the animals and plants by which it has been inhabited, have undergone a great succession of changes. In the rest of Europe, generally, the volcanic rocks have either been originally of submarine origin, or the surface, since they were produced, has suffered so much denudation by the action of the waves of the ocean as to make it impossible for us to ascertain the form and manner in which the eruptions took place, or the relative position which the igneous formations held at first to the hills, plains, and valleys then existing. After describing the several classes of rocks in Auvergne, -the granite, the eocene fresh water, and the older and modern volcanic, each depicted by different colours in an extensive landscape, enlarged from a view of the valley of Chambon (Puy de Dome), by Mr P. Scrope, Mr Lyell said he should dwell chiefly on the antiquity to be ascribed to the Puy de Tartaret, a type of one of the most modern cones of eruption in Central France. The compara

tively recent origin of this conical hill of scoriæ, with its crater at the summit, is proved by its standing at the bottom of a deep valley excavated through the alternating beds of pumice, trachyte, and basalt, belonging to the more ancient volcano of Mont Dor, and partly through the subjacent and fundamental granite. It is farther confirmed by the course of a powerful current of lava; which, proceeding from the base of the cone, flows thirteen miles down the channel of the river Couze, stopping at the town of Nethers, near Issoire. The lava occupies the ancient river bed, and is observed to contract in its dimensions in the narrow gorges, where it also gains in height, like the water of a river flowing through the arch of a bridge; and to expand again where the valley opens, where it speads into a broad sheet having a level surface. It also flows up the channels of tributary streams till it attains a level corresponding with the top of the lava at the point of junction of the tributary with the main valley. But although these appearances prove that the lava has flowed as it would now do if it were remelted and made again to descend the

* Account of a Lecture read before the Royal Institution on April 20, 1847, by Mr Lyell.-Athenæum.