in his plan and comparative views are given of the different changes which they undergo, whether in the Teutonic forms of speech or in others of a cognate origin. This is one of the most important and elaborate portions of the work, and that, perhaps, of which the value could least be supplied from other sources. We shall, in a subsequent part of this paper, take occasion to notice some of its results in relation to the comparative structure of the classical and Teutonic languages. With respect to the Teutonic languages themselves, it may be observed that, except in the case of the High-German dialects, to be afterwards mentioned, the consonants, for the most part, exhibit comparatively little change in passing from one language to another: but the vowels are on a different footing. This liquid and fluctuating element of speech seems to assume a new aspect in every new situation and may be said to furnish the great feature of individual distinction between different dialects. The changes, however, to which the vowels are subjected, seem capable of being reduced to determinate rules, and these it is obviously of great moment to ascertain. We would not argue that the clear and certain principles of any language are to be controlled by mere theory derived from others: but it seems an indisputable proposition, that what is dark or doubtful in one of a family of dialects, may be illustrated and rendered plain by the light of a regular analogy obtained from kindred sources. An attention to this enquiry is also necessary, if we wish to trace the connexion between the words of one language and those of another, as the disguises assumed in the process of transition are often such as to elude or deceive a superficial observer. The following are examples of the changes which the vowels undergo in passing through some of the chief Teutonic tongues. The convertible sounds are given in the same horizontal line.

Gothic. O. High-Germ. A. Saxon. Icelandic. air ir, ēr

iar

or

eor

ei

ject, which must be afterwards adverted to.

The Second Book treats of the inflexion of words. A short experience enables us to see, that the modern forms of inflexion, limited as they are, cannot be understood without an acquaintance with the earlier types on which they have been moulded, but of which the features are now often greatly, and somewhat unequally obliterated. We presume it will be universally agreed, that in this part of his subject, and in so far, particularly, as concerns the declensions and conjugations of the Gothic language, the Grammar of Grimm has rendered a great service to philology. As compared with any former grammar, and especially with any of English production, the progress in sound and systematic arrangement is incalculable. Particular importance is here due to the formulæ of the twelve strong or primitive conjugations; that is, of those conjugations which are effected by a reduplicative prefix, or by an internal change of vowel, instead of receiving the addition of an element corresponding to the ed of the preterite and past-participle in English.

It was long ago observed by Ben Jonson, in his Grammar, that, while that conjugation in English which takes place by means of the suffix ed, was "the most useful forming of a verb, and thereby also the common inn to lodge every strange and foreign guest," the other mode of conjugation, by means of an internal change of vowel, as fall, fell; break, broke; sing, sang, sung, &c., "entertaineth none but natural and home-born words, which, though in number they be not many, a hundred and twenty or thereabouts, yet, in variation, are so divers and uncertain, that they need much the stamp of some good logic to beat them into proportion. We have set down," he continues, "that that in our judgment agreeth best with reason and good order, which, notwithstanding, if it seem to any to be too rough-hewed, let him plane it out more smoothly, and I shall not only not envy it, but, in the behalf of my country, most heartily thank him for so great a benefit, hoping that I shall be thought sufficiently to have done my part, if, in tolling this bell, I may draw others to a deeper consideration of the matter; for, touching myself, I

Grimm's Grammar on the theory of the Teutonic conjugations, embody a list of 460 strong or primitive verbs, with their comparative forms in the different dialects, which deserve to be familiarly known to all Teutonic philologers, as collecting together the most valuable materials for etymological illustration.

The Third Book is devoted to the formation of words by derivation and composition. The subject is as fully as it is ably discussed, and occupies two volumes, containing ten chapters. The first chapter explains the derivation of cognate words from the various conjugational parts of the strong verbs. A very large mass of vocables from all the Teutonic languages are here brought together, and the commen bond of their connexion developed by Grimm-a task which had before been laboriously and usefully, but somewhat loosely, performed in the Dutch work of Lambert Ten Kate. * The suggestions made by Grimm as to the possible forms of strong verbs now lost, but of which the disjecta membra survive in their derivatives, may occasion differences of opinion among philologers, but are at least highly valuable as illustrations of the limits within which the affiliation of words may with certainty range. The second chapter treats more peculiarly of derivation, and digests under distinct heads the various terminational portions of words by which the root is lengthened in form and modified in signification. The third chapter treats of the composition of words in all its varieties, and illustrates the subject with copious examples. The fourth treats of the formation of pronouns. The fifth of that of adverbs, prepositions, and particles generally. The sixth embraces the difficult subject of gender, which is discussed with more fulness and systematic arrangement than, we think, have before been bestowed upon it; while the illustrations given may be said to furnish in themselves a rich treasure of information on many points of ancient Teutonic manners and modes of thinking, as well as of terminology. The seventh chapter treats

of comparison. The eighth of diminution, or the formation of diminutives; the ninth of the forms of negative words, and the tenth of those expressing question and answer.

The Fourth Book is devoted to syntax, a subject which the author has not yet completed, or rather has scarcely commenced; his fourth volume, which is the latest published, being confined to the syntax of the shortest and simplest models of a complete sentence. The more complex forms of phraseo. logy and composition are reserved for the continuation of the work.

It is not our intention, if it were in our power, to raise or revive any question as to the accuracy of all the views adopted by Grimm, in reference to individual languages. His great work, minute and detailed as for the most part it is, must yet be considered chiefly as a Comparative Grammar, in which it cannot be expected that in each particular department the slighter features should be as strongly delineated as the more considerable; or that the same clearness of perception or accuracy of knowledge should be exhibited, as if it consisted of different and divided dissertations on its several topics. A geographical treatise may be allowed sometimes to be loose or incorrect as to the topography or statistics of a county or a parish, and its oversights may be detected by microscopical observers who are incapable of reaching or of understanding its comprehensive results. The Teutonic Grammar is naturally fuller and more perfect in some departments than in others: as, for instance, in the Gothic and in the Old and Middle High-German. But in all its parts, so far as we can judge or learn, it is more minutely correct than most of the existing individual grammars; while with that minuteness it combines in an unparalleled degree the advantages of historical deduction and varied illustration.

It is manifest that a complete survey of the Teutonic languages, in the spirit which we have attempted to explain, must be of the utmost service to philology. It has been fully established, what was at one time but imperfectly understood, that the whole of the languages thus considered are

originally identical, not only in their radical words, but in their structure and inflections. It becomes apparent, at the same time, not only that it is impossible scientifically to understand any modern Teutonic language without a knowledge of its more ancient form, but that it is equally impossible to know thoroughly the character of any one branch of the common stock without a reference to the correspond. ing features of its sister shoots. Above all, we may be allowed to say, a scientific acquaintance with the Gothic language is essential to a thorough knowledge of any of the other varieties of Teutonic speech, whether ancient or modern.

The importance of a comparative study of the Teutonic languages, in a manner at once comprehensive and minute, must be peculiarly apparent with reference to all enquiries of an etymological nature. It will often happen that the root of an extensive class of words is only to be found in one or in a small number of the Teutonic languages, while its derivatives are widely diffused, or exclusively confined to others of the family. We shall sometimes, again, discover the primitive or proper meaning of a word in one tongue, while in others it has been diverted to a merely secondary or figurative signification. A wide view of the whole field will at once conduct us to discovery, and guard us against those errors of precipitate and presumptuous dogmatism, which have often made the school of Horne Tooke a legitimate laughing-stock. In endeavouring, for example, to analyse a supposed derivative word in any one language, we may be rashly led to imagine resemblances to simple elements, whether in form or meaning, which a larger examination of the affiliated dialects would prove to be erroneous. If in two or more Teutonic languages, of equivalent authority, we have the same word which we propose to explain, and if in those languages we have also the elementary words of which we conceive it to be a product, we ought not to infer that our supposed etymology is correct, unless in all of these languages our conjecture be equally consistent with their forms and rules of derivation. We may point out one or two common fallacies in etymology, which we think have in this manner been detected.

It is a frequent and a flattering remark that the English name of the Deity, God, is a derivative from "good," and must have been selected by our ancestors as indicating the best and loveliest of the Divine attributes. Junius,a writer whose own learning and ingenuity have greatly assisted his successors in correcting his numerous errors, has, in his Etymologicum Anglicum, expressed himself thus under the word God::-"Origo vocabuli dilucidè satis desumpta est ex A. S. god, bonus. Nam ut sapientissimi Teutonicæ linguæ authores, hominem, inexhaustum omnium vitiorum turpitudinumque gurgitem, man dixerunt, a man, malitia, &c., ita Deum, perennem bonitatis æternæ fontem, antithesi valde manifestâ atque illustri God dixerunt, a god, bonus." A similar derivation and remark is to be found in Mr Bosworth's very useful Anglo-Saxon Dictionary. But both of the etymologies here referred to are shown to be more than questionable, by reference to the same words in other lan guages. It is only, however, as to one of them that we shall here point out the fallacy. It is true that God, Deus, and god, bonus, are spelled in the same way in Anglo-Saxon, though the prosody is different; but a wider examination shows that the words are es sentially distinct: the radical vowel in the name of the Deity being a short u, while in the adjective for good it is a long or double a. There is no principle by which, in the Teutonic languages, the vowels, thus diversified, can warrantably be referred to the same roof. The true etymology of the Divine name is still a dark and disputed question.

Again, to take another and humbler instance :-Acorn, in Johnson's Dictionary, and even in more modern works, is described as being compounded of Ac, Anglo-Saxon for oak, (aik, Scot.,) and corn or cern, as AngloSaxon for a grain or nut. This is violent enough, even upon Anglo-Saxon principles; but the theory is discredited by the existence of the word in other languages, in such a form as to throw the utmost doubt on the identity of the alleged compound with its component parts. Thus, in Icelandic,. Akarn is the word answering to acorn; but the Icelandic Akarn cannot with any propriety be derived from its supposed elements, of which the forms

in that language are eik, quercus, and korn, granum. The Gothic Akran, glans, fructus, is most probably the original of acorn, and is certainly derived from Akrs, ager. It thus means literally the produce of the earth, and has only become applicable to the fruit of the oak, as being a distinguished article of diet in that stage of society, "When wild in woods the noble savage ran."

"Glandis appellatione," say the civilians," omnis fructus contine tur." It is possible, however, that, in some of the dialects at least, the term may have come to be more peculiarly applied to the fruit of the oak, from a vague feeling of its resemblance to the name of that tree; a process which we believe to operate insensibly in the use of language to a considerable

extent.

The multiplication of checks and assistances of this kind seems essential to the due prosecution of philology, whether we contemplate the wide field of discovery which it opens to us, or remember the snares and pitfalls to which our progress in it is exposed.

The first elements of language, like the primitive atoms of matter, lie hid and disguised in their actual existence under innumerable varieties of form and combination, eluding the eye of superficial observation: and it is only by systematic study that we are even partially enabled to reproduce them in a more simple state, and to discover, on the one hand, the real identity of objects that appear different, and the real diversity of others that appear the same. As the original sources of human speech are hid in the darkest antiquity, it is conceivable that all, or many of the languages with which we are acquainted, though their courses and channels be now far distant, or mutually opposed, may yet have flowed from fountains springing up in the same spot or vicinity. The inhabitants of Leyden and Lyons may unconsciously drink the snows of almost the same glacier; and the German or Scottish peasant never dreams that the unintelligible dialects of the stranger or scholar are, often mere varieties of his vernacular tongue. If we consider the infinite combinations of which elementary sounds are susceptible, and the infinite changes to which they may be subjected by the attrition of use, the

corruptions of rudeness, and the refinements of affectation, we can have no difficulty in understanding how languages, diverging from a common centre, should in time lose any obvious trace of mutual affinity. The instance given by Dr Watts of the apparent dissimilarity and real identity of the words bishop and evêque affords a good measure of the divergences that have taken place in the diffusion of the Roman language among different countries since the establishment of Christianity. Another and less familiar example may be given, which reaches into a higher period of antiquity.

There is no doubt that the English word tear, and the French word larme, are the same in etymological origin as well as in meaning. Tear is referable, through the Anglo-Saxon tæher, to the Gothic tagrs, which is of similar meaning, while tagrs is demonstrably identical with the Greek dazgv. Aangu, however, or its derivative and synonyme dangupa, is, without doubt, another form of the Latin lacryma, through the interchange, which not unfrequently occurs, of the letters land d. Finally, lacryma, which is thus traced to be a cognate of tear, is the acknowledged parent of the French larme. A connexion is here established, by a plain and necessary procession of proofs, between two words which have no external resemblance whatever. The steps of the deduction appear simple when they are shown; but if we were unacquainted with the Greek and Gothic links of the chain, we should never suspect the intrinsic relation of the extreme points. To a great extent we are placed in that supposed state of ignorance as to the common bonds by which diversities of language are reconciled; an ignorance which in the case put would have left room for boundless conjecture, instead of the direct demonstration afforded by a fuller knowledge.

The latitude of speculation which is opened up by these peculiarities of the science constitutes one danger, as well as one attraction, of etymological pursuits. Because infinite analogies and affinities may be conceived among remote forms of speech, the etymologist often hastily and falsely infers that he has discovered their existence. The interest and the extent of the subject heat the imagination of the enquirer,