and, like the day-dreamer gazing on the clouds, or on the embers of his winter fire, he is prepared to trace the most fanciful and unsubstantial resemblances, where the sober eye of reason can make no such discovery. Occasional results, too, of real value lure him on, like the alchymist, in the search of some universal solvent of all difficulties, that has either no actual or no relative existence, and only leads to absurdities that bring the study into disrepute.

It is only by treating philology as a proper science that its reputation can be sustained, or its progress promoted. If it is at all a fit subject of accurate investigation, it must be dealt with, like any other branch of inductive knowledge, by the twofold process of collecting, extensively and minutely, the facts connected with it, and of endeavouring to generalize those facts into distinct laws. The facts collected ought undoubtedly to be extensive, at least within the particular province that may be selected for examination, otherwise no comprehensive or safe results can be obtained. But it is the spirit of generalization that can alone reduce the chaos of our materials into order, or animate them with life. We have no reason to believe that any of the phenomena of nature are governed by a lawless caprice. The heavenly bodies do not roll on high without a fixed decree to impel and restrain them. The wildest brook that wanders across the plains obeys, in all its windings, the dictates of a steadfast rule. We are not to infer that language, one of the noblest gifts of God, is left to shift and fluctuate from tribe to tribe, and from generation to generation, without regular and constant principles of change. In this, as in all other departments of observation, we may rejoice to address ourselves in the language of the poet:

"All nature is but art unknown to thee: All chance, direction which thou canst

not see.

And this further is permitted us to believe, that although the ultimate secrets of nature must be for ever hid from mortals, yet a partial and pro

gressive discovery of her laws is promised to us by past experience, so long as the study of her works shall be prosecuted in a humble and hopeful spirit, and with a firm reliance on the wisdom and uniformity of her operations.

The field of enquiry which chiefly seems, in our own day, to call for the labours of the philologer, is at once extensive and captivating, and is becoming daily better defined and understood. It comprehends that vast family of languages which, reaching from India to Iceland, has embraced within its bosom almost the whole range of human civilisation and uninspired literature. The Indo-Germanic tongues present a subject of study of the most sublime and inexhaustible interest. Their original identity, not merely in their radical words, but in their whole system of inflections, appears to have been fully demonstrated. What has been so thoroughly accomplished for the Teutonic languages by Grimm, is now in the course of being still further extended by Bopp, in his "Comparative Grammar of the Sanscrit, Zend, Greek, Latin, Lithuanian, Old Selavonic, Gothic, and German ;"* a work of profound learning and consummate ability, of which, so far as we can judge of it with a very limited knowledge of its subjects, the execution appears to be answerable to the promise which its title puts forth. It remains to follow out the same principles into the apparent dissimilarities of Celtic philology, an object which, until lately, has been too much neglected by the scientific students of comparative philology; but which is now, we believe, begun to be widely considered in its true light. Although compelled to acknowledge our ignorance of this important branch of the subject, we can see evidence that in our own country we have both classical and modern scholars among us who are able to contribute valuable assistance in this department; and the labours of Williams and of Pritchard must remove the idea, hitherto prevailing, that where Celtic learning was concerned, the amare et sapere" were irreconcilable things. †

66

In a comparative study of the numerous household of affiliated languages, to which we have now referred, it is of the utmost importance to reduce, if possible, to definite laws the principles of transition which prevail among them. Nothing can be more unsatisfactory than the system, which so long existed, of inferring resemblances from mere external appearance and general impressions, without a minute deduction of the historical etymology of words, under the separate heads of structure and signification. The Grammar of Grimmought to introduce a new era in studies of this description. It has placed it beyond a doubt that a fixed principle prevails in comparative philology, where every thing was formerly thought to be arbitrary and accidental. It has contributed to shed a clear and steady light on a dark domain of science, where before we had nothing to guide us but an occasional Will of the Wisp, that almost always led us out of the right road, and generally landed us in a quagmire.

Etymology, as already hinted, may be considered as consisting of two branches, which ought ever to go hand in hand; the physical and the metaphysical:-the one treating of sounds, the other of ideas, in their several relations and susceptibilities of variation. Each of these divisions presents a large field of enquiry, and in each there is ample room for the application of philosophical principle. Let us take, among many illustrations, a remarkable example of a natural affinity subsisting between different ideas, and leading to a modification of the meaning of cognate words, which has had an important influence on language.

Sound and light, the objects of the kindred senses of hearing and seeing, are bound together by strong mental analogies, and the same radical word is often, in the same or in cognate languages, employed to signify those several conceptions. To give forth sound, and to give forth light, to speak and to shine, though often very different things in the House of Commons, are in Greek and in Sanscrit denoted severally by the same root. Bhami the Sanscrit verb signifying to shine, and paμ, the Doric form of

φα-ντι.

bhâ-tha bhâ-nti

Donaldson's Cratylus, p. 558. Clarus, clear, is, in like manner, an epithet equally applicable to the objects of hearing and of vision. The English loud, representing intensity of sound, is very probably a relative of the German lauter, signifying bright, pellucid; and both words appear to be connected with clarus. Dim and dumb seem to be cognate terms, indicating the obscuration of sound or colour; and on the same principle there is reason to connect together the Latin surdus and the Teutonic swart, which, though differently applied, refer severally to the absence or negation of those kindred qualities.*

With respect to the structural part of etymology, it appears to be now demonstrated that a very singular and settled relation subsists between the chief tribes of the Indo-Germanic race of languages. The liquids, sibilants, and semi-vowels, remain generally unchanged in them all, under certain known modifications in individual cases, such as the frequent substitute in Greek of the simple aspirate for the ordinary sibilant, and the ultimate disappearance or modification of the w or digamma in the same language. But the history and position of the mute consonants in those tongues, is more peculiar. The general rule as to these is, that, in the transmission through different languages of words characterised by those consonants, they undergo a certain fixed and regular modification. We are not, of course, now speaking as to any change undergone by words imported directly from one language into another, such as the Latin, Greek, or Norman terms, which abound in ordinary English speech. We refer to those more venerable vocables which appear as natives in a great variety of languages of different

character, but connected by some ancient bond of consanguinity. In words of this description, it seems, as we have said, to be demonstrated that the mute consonants composing them are not to be found in the same form

throughout the different languages, but are subjected to certain important permutations, according to a definite rule.

The rule referred to, in its highest theoretical perfection, may be thus broadly stated, leaving out of view those qualifications and restrictions upon it, of which a detail can only be expected in a systematic treatise. Viewing the languages of Greece and Rome as in this respect on the same level, and contrasting them, as one branch, with the Gothic or Low-German dialects on the one hand, and with the Old High-German on the other,

those three sections of the Indo-Germanic race are to be considered as occupying three equidistant points in a supposed circle. The mute con. sonants, again, being divisible into three well-known orders, the tenuis, (,,,) aspirate, (x, 9, 4,) and middle, (v, d, ß,) and these being supposed to move in that order within the circle containing the three divisions of Greek, Gothic, and High-German, an index will thence be obtained to denote the modifications which these consonants present in those different languages. Thus, where the same root exists vernacularly in all the three great dialects, the consonant which, in Greek, is a tenuis, will be found, in Gothic, as the corresponding aspirate, and, in passing on to the High-German, will become a middle. The Greek aspirate becomes a Gothic middle and a High-German tenuis. The Greek middle a Gothic tenuis and a High-German aspirate.

This statement of the rule will scarcely be intelligible without a visible figure. It may be put in another

way, and will perhaps be best under.. stood from the following formula.

Greek-Gothic-H.-German.

Aspirate- Middle-Tenuis-Aspirate-Middle.

Supposing these parallel lines to be scales, of which the upper one is horizontally moveable towards the left: the fixing of any one of the languages opposite to any of the order of consonants, will show the corresponding change which that consonant exhibits in becoming naturalized in the two other languages.

It must be observed, that in the application of this rule to different languages, there are several accidental anomalies which disturb its operation. Thus the Latin is without the dental

aspirate, and generally supplies the place of the guttural aspirate by a simple h. The High-German is, in like manner, deficient in the dental aspirate, and supplies its place by a ts or z. The High-German is subject to one or two other irregularities, which it would here be out of place to detail.

We shall now give examples, in the dental class of consonants, of the

changes we have described; which, it must be observed, are not reciprocated between any two languages, but revolve through the whole three. The following examples, to the other classes rule may be easily extended, from the of mute consonants.

=

NO. CCXCII. VOL. XLVII.

The system of transition now noticed has commonly been called Grimm's law of "Lautverschiebung," or Sound-shifting, as it may be loosely rendered. But it is due to the memory of a very great philologist to observe, that it was anticipated, in a partial but most important point, by a previous writer. The mutual relation of the Greek and Scandinavian, in their initial consonants, had been formally announced by Erasmus Rask, in his "Enquiry into the Origin of the old Northern or Icelandic Language," a work which was published in Danish in 1818, and of which some of the tabular views were subsequently translated by Vater, and included in his Vergleichungs Tafeln." The first edition of Grimm's Grammar appeared in 1819, and it was only in the second edition of 1822, that the general law referred to seems to have been properly developed. Rask did not, perhaps, see the full extent of his own discovery, particularly as to noninitial consonants: though it must be confessed, that as to these its operation is sometimes considerably disturbed, and it would seem that some peculiarities in the Scandinavian tongues might, in this respect, tend to obscure his perception of the rule. Grimm, on the other hand, was probably enabled to follow it out more confidently, from being led to observe that the old High-German stood in the same relation to the Gothic and its dependents as these to the Greek or Latin.*

In the older philologers we find little or no indication of the law now

adverted to. It is repeatedly, indeed, observed, that mutual transitions take place among the tenues, middles, and aspirates of the same class of consonants; and it was impossible that the frequent occurrence of such changes could escape the most cursory enquirer. But they are treated by those writers as the exceptions and not as the rule; they are as often misapplied and mistaken as the reverse; and they are seldom resorted to for explanation till a more direct and literal etymology, however desperate, has been attempted in vain. Thus Junius, instead of connecting the Gothic dauthus, death, with Savaτos, which seems its most probable etymology, derives it from the Greek day, longævus, and gratuitously compliments the Gothic nation on their lofty creed, thus philologically promulgated, that death and immortality are the same things.

It is a just observation of Grimm's, that, according to the law which we have attempted to explain, a correspondence between the mute consonants of words in different sections of these affiliated languages is, generally, a proof, not that the words are the same, but that they are different in origin. We do not indeed affirm that the law is universal, and without exception in its operation. Some words of hardier fabric, or of happier destiny, have undoubtedly floated down the stream of ages, untouched by the influences that have disguised or mutitilated others. We are told that the word sack is to be found in the same form in almost all languages, which gave occasion to the facetious observa

*

tion of the learned Goropius Becanus,* a gentleman whose lucubrations we are little acquainted with, except in connexion with this jest, that at the building of Babel," nemo ædificantium in subitâ linguarum confusione oblitus est sui SACCI. Other vocables, and parts of vocables, have made a similar escape from the effects of time or transition; but such exceptions would not disprove the existence of the general rule, even if they were more numerous and unequivocal than we believe them to be. Further enquiries, we are inclined to think, will show that many apparent cases of affinity, where there is an absence of that change of consonants which the rule would require, are not real deviations from it, and that thus a number of common and plausible etymologies are unfounded. There is little doubt that the Greek paulos, and the Teutonic foul, are not connected in etymology: as indeed they do not appear to be in origina! meaning. The Latin Dies has perhaps less connexion with the Teutonic day than with the terms time or tide, to which, according to Grimm's law, it may radically cor respond. The derivation of care from cura, is doubtful, and the common identification of vulgus and folk, if not unfounded, is, at least, remote and indirect.

We have dwelt so long on this subject, from a conviction that the full truth and practical importance of this singular and mysterious law of transition is as yet but imperfectly felt in English philology. Within the last few years numerous works in lexicography and etymology have appeared in this country, which continue to be constructed on the same principles as if Rask had never lived, or as if the Deutsche Grammatik had never been written. Thus, in a Greek lexieon of no. remote date, and in other respects valuable, we find such English etymologies as the follow

ing-ançodgva, acorns; fλur, blast; κνημη, knee ; κοίλος, to coil up ; πολλοψ, collops; duga, dirge, &c.; and worse examples might be derived from other sources. It must, on the other hand, be confessed, that ordinary Teutonic philologers often lay themselves open to the charge of rashness or ignorance when they venture far out in the sea of classical philology. It must rarely, indeed, happen that a thorough acquaintance with each department is united in the same individual and perhaps in no branch of science is partial error or occasional oversight more probable or more pardonable.

One or two recent publications will greatly tend, we think, more widely to diffuse a due knowledge and appreciation of Grimm's rules of comparative philology. Mr Pritchard's work on the "Eastern Origin of the Celtic Nations," (1831,) has attracted the attention that is due to any production from so valuable a source. "A Manual of Comparative Philology," by Mr Winning of Bedford, (1838,) is entitled to much commendation, and deserves to be generally consulted, though we do not avow ourselves converts to his theory on the origin of the Tus cans. Of the New Cratylus of Mr Donaldson, (1839,) we must not, as yet, profess to have formed any other opinion than that it is a work of great learning, of much interest and value, comprising a mass of materials that have never before been collected in an English shape, and proceeding, in the main, on what we perceive to be sound and safe principles, though sometimes, we humbly suspect, extending into a latitude of speculation that is at least premature. In all of these works the views of Grimm are explained and enforced.

Having said so much of the relation subsisting between the consonants in the different sections of the Indo-Teu