A calm consideration of what is meant by a relative clause will perhaps clear this anomaly of all substantial difficulty. A sentence which involve a relative consists of two clauses ; that which contains the relative is used solely for defining something, which is to be 'Token of in the second clause ; and it Is this second clause alone which carries with it the news that the speaker wishes to convey. If the object to be defined be present, be may point to It, and no further definition ifs required. If otherwise, it is convenient to suppose it for the moment to exist in the form of anything at hand. Thus, the lawyers of Rome, in conveying a distant estate in land, took up any clod of earth, and treated it as the actual land then selling. Thus one might say, " This bone I bought yesterday,—this horse I will eel) yon for 201.;* or in Latin, " Hone (or quem, no matter which) equtun heti emi, hunt tibi vondama Then, again, identity for the most part in form of the relative and interrogative needs not a word in proof, and the identity in power is clear in the case of the indirect interrogative. Thus, " Dic gni. feat' (Arena in old writers is not necessary), might be translated by Show me the man who did it," saving that an s happens not to be used with the relative; but this s has nothing to do with the base of the word, and indeed may at time be omitted from the interro gative itself. To translate die by " show ," is simply to revert to the older meaning of the word ; and between the two phrases eels feel: I and die gnit fecit„ to give a precedence in time to the latter, is in accordance with the habit of language, which ever colts down long phrases to shorter ; and in this particular case the omission of die was the more natural as it was the standing commencement of every ques tion, and therefore wholly superfluous, if the manner of the speaker suggeeted that he was asking a question. At the dinner table, "A glass of wine / " is quite intelligible without any interrogative whatever.
As the meanings, then, run all into one another, the chief question remaining is as to identity of form. Now it will be found that an initial thin guttural is found in qui and quiz, in the Creek NOV, ?forgot, in our own old pathos, ger/acre. Theo, again, for the demonstrative, we have in the Latin cis, dire, eitcrior, &c. ; in the Greek KFIVOI, or wenn ; in the Tuscan quells, questa, cold, coati : and in a question of this kind. the Tuscan stands on equal ground with the Latin, for we know that the language of Tuscany, even in the days of Roman supre macy, abounded in gutturals. As for the other consonants which are seen at the oommencement of third person pronouns, all surprise at the change must fall sway when we look to woe supplanting KOV in Greek, pipit supplanting quiequid in Latin, while the Latin vas is the ana logue of the Creek en, just as que," and," is of ye," and." And if an s la even In the Latin sir, k:ng. so and such, and the compound whoso, we have what is perfectly parallel in the Greek 0117ft, naprpov, for Tern, "viper. Then, again, the mere aspirate of 6, beside so, and of et, 41. 4, of the Latin hie, and our own he, here, hence, is just the latter which is constantly superseding, now an s, now a Is Or if the loss of the militate Itself in is, ra, id, be premed, we have only to call to mind that what in now written it by ourselves was formerly written /tit, and is still Ad in Dutch. A guttural is of all letters most likely to be corrupted, and hence, If there be reason to suspect that the third pronouns were of one origin, as in sense they might well have been, the k has a stronger claim by far than any of the others, p. t, s, A, or torn. se to the end of the word, it will be found that on overlaid. a meal prvente itself. The Greek exhibits such a letter in vne (n. en) ; the Latin in use-dc, atieeptie ; the Spanish in quirn ; the foregoes,. in gate ; the Swedish in a nom. /went, gen. harems; the Dutch in a gen. tit1/1/4 ; the English in whooce (old whcanso). The Sanskrit grammarians, again, insist that the base of their relative Is lies, not li ; and beyond the Iiiiiite of our family, as usually defined, the Finn lam Lou fur the relative as well as re : and the Turkish Lim, se well as Ish and Li. Then. again, for the forms which prefer an s, the Norse relative Is arm ; the Daniell sons, still retained in our own despised Ameseesoner. The Slavonic ferias for "this" or "that" are ten and sew in Bohemian, to in Riusian. The None, for its personal a masculine, has N. Aetna, ac. Aosta, gen. AO/1/1, dat. heat Pi; and article the same languae has • final a In le nom. m. f. re, and carries this w Into every oblique of all genders and both numbers. With this evidence so strongly converging to one point,— mostly, the ayllable Len—the question is, what did this ken originally denote L One cannot but answer that our own obsolete ten in the seam loek" ',applies that meaning of all others which Is beat fitted to be the enigin of demonstrative pronouns. The one difficulty ix, can
we establieh the antiquity of this verb I Something has been already wad on title point. It lam been traced to the Latin and Greek se the root of gnoseo, gnori, °gait us, coyititais. But it also occurs as a particle in Latin, but of course in a reduced &hope ; for when verbs are degraded to such an office they are always reduced, as in our own lo, for look. The particles alluded to are en, "behold,' and the suffix ce, attached exclusively, as it well might be, to demonstrativea, as to hi-c, lard-&, sit, tun-e, or, in fuller form, hun-ce, Ai-ei-ne. Lastly, far the antiquity and wide extent of the verb ken, we need but add, that It elide in Chinese with the very sense desired. It has been thought right to enter into these details in order to establish a point so utterly opposed to. the doctrine of pronominal roots as taught by Bopp, and too readily adopted by his followers ; and by way of enlivening a subject which to many may appear dull, it may be noticed that occasionally the force of a Latin sentence Is brought out more vividly by recurring to the primary meaning of hie and Me. Thus, in Terence, ' Haut.' 8, ], I, Leteiseit hoc ions--" It is getting light, look, already ; " and in ' Virg.', 5, 457, Pracipitenique Dorm ardent agit ortore toto, Nurse de.rtra ingrminans ietus, nurse We sinistra—" Now with his right redoubling blows, now, look, look, with his left." The classification of the languages of the world Is as yet in a very imperfect condition, as might be expected from the fact that a very small portion of them have yet come fairly within the reach of science. Of the great majority, we possess but short vocabularies, and those obtained tinder circumstances which justify a strong doubt of their value. It is therefore a matter of but ordinary prudence to abstain from all definite theories as to languages of which so little is known. The Indo-European family is in a very different position. Belonging for the most part to highly civilised societies, each member of this family has at least a dictionary and grammar; and having been long under the scrutiny of men of science, they have severally yielded results of value. Yet even hero a too hasty generalisation has been made. It was long before the Keltio branch was admitted to its full rights as a member; and philologers of the German school still draw a line of absolute division between this family and other languages, the accuracy of which the unfettered inquirer must bo permitted to doubt. The Finn and Lapp family, for example, are found to have affinities with the Lido-European stock, which are not to be explained by their being so closely in contact with the races, for over and above those importations which are sure to take place where barbarism borders on civilisation, there aro coincidences which this con tiguity will not explain, such for example as between the Lapp mecum, tot-um, sot-um, and the equivalent forme in the Latin mecum, lectern, serum, for pronouns and simple prepositions are precisely the sort of articles which are never imported. Again in Nome the negative is el, eigi, in Danish itke ; and in Finn ei, eiL-17. The personal endings of a Lapp verb also bear a close resemblance to those of our Indo-European family. Thus the plural of the verb motto " change " runs, raolsoinien, molsoite, niolsoin. The interrogative and relative in Finn have gene rally the form cu, with suitable affixes for the cases, &c. Again the verb lasirn," let go," and its diminutival form laskelen, "let go by little and little," exhibit both in the main syllable and in the muftis el, what is strikingly like words in the great European family. Lastly, as the Latin by adding the particle que to the relative creates quisque, so in Finn the addition of a syllable not unlike que to the relative jo, that is, yo (so like the Sanskrit ye) produces joca, one." In truth a community of pronouns with the Indo-Teutonie family may justly be claimed for several languages which lie far outside the limits as yet assigned to it. To hint at a connection between our family and the Semitic, would shock the ears of most German philologers ; and it must be admitted that a sad amount of rubbish in this way affords some excuse for the prevaleut dogmatism ; yet the two Hebrew numerals shish and shih' for " oix " and "seven," raise, to say the least, a strong suspicion of some affinity, though, no doubt, it must be a most distant affinity. Be this as it may, the Chinese exhibits so many evi dences of identity in important words with the Indo-European family, two examples of which have been here given, that Porno relationship must assuredly exist; and to connect two such remote families is virtually to draw into the circle many others.