adopt Grimm's view of the nature of the particle how, to be reasoning in a circle, in deriving the relative prefix w from a case of the very word who, which it is our object to explain.

The Abbé Sicard, in his Elémens de Grammaire Générale,' appears to have come nearer these views than any other author we have seen. He was placed in circumstances peculiarly favourable to such researches, by the office of Director of the Schools for the Deaf and Dumb, which he filled with so much distinction, and where, in order to teach language to those who could have no conception of its meaning, it was obviously of the utmost importance to reduce all the principles of grammar to their primitive elements, wherever that could be done, but always to the most simple and intelligible possible. All elliptical and abbreviated modes of expression would require to be filled out, and expressed in terms of their component ideas, before he could hope to make them understood by the scholars, with whose minds he had such inadequate means of communicating; and he had thus, in the comparative facility or difficulty of effecting this purpose, a touchstone, as it were, on which to try the truth of his analysis. In speaking of the relatives, the Abbé says,- These words, qui and que, are elliptic expressions which answer, in their first part, to the letter in algebra, the unknown quantity which replaces the preceding term, or points out that which follows, whether this term is a noun or an entire proposition; and which are, in their second part, either the pronoun il in qui, or the verb est in que.' If it had not been for the unlucky explanation of the e in que, it might be said that he had here touched upon the very verge of the true meaning of these words; but it only requires to cast our eye over his illustration to be convinced of the worthlessness, in grammatical studies, of any but the most well-defined and exact notions. The following is the example given by M. Sicard of the use of the relatives, together with his mode of decomposition :

L'astre que nous admirons le plus est aussi celui qui nous est le plus utile.

Nous admirons le plus qu;

Est cet astre est qu;

Il nous est le plus utile.'

In this decomposition it is easy to see which are the words replaced. by the two qu's. We also see the e of que represented by est, and

the i of qui by il.

The remaining words, of a relative nature, that we possess in English will give little trouble, as they are all immediately derived from the pronoun who, of which, indeed, the adverbs when, where, and why, seem to be simply cases. It will perhaps illustrate this

view of their nature and derivation, if we give the declensions of the relative and demonstrative pronouns in Anglo-Saxon:-

It will readily be seen from this table how near then and when (in Anglo-Saxon, thone, hwone) come to the accusative masculine of the relative and demonstrative pronouns respectively. There seems to be the genitive feminine, there-in analogy with the Greek TOU, which is also of a genitive form, although it is in the masculine gender. The relative pronoun has no feminine in AngloSaxon; but as it is formed in all existing inflections in exact accordance with the demonstrative, the genitive feminine, if there was one, would be whære, and it is probably from this source that our where is derived. Why is the ablative essentially unaltered; corresponding to which we had formerly a thi; forthi being of constant recurrence in old English in the sense of on that account. There was thus originally nothing in the word when which conveyed the idea of time, nor in where of place; but these being the cases of the relative pronoun by which reference was made to the time when, or place where, any action occurred, the usual tendency of language to abbreviation showed itself in the omission of the nouns of time and place. The pronouns when and where would then seem to contain in themselves the meaning of the nouns respectively understood before them, aud would thus put on their present adverbial appearance. In the very expression we have just used, of the time when any action occurred,' the meaning of when is even at this day simply pronominal, and we might with equal propriety have said, the time at which any action occurred; and the pronominal nature of these adverbs is still more apparent from the use of where in composition,-whereby and wherefore meaning by what and for what, without even a reference to the idea of place. Which, in old English whilk, was originally who like, and was identical with the Latin qualis (Gothic, hueleiks), although it has in modern English, in certain cases, usurped the place of the simple pronoun what. On exactly the same plan are formed the Scotch ilk, the same, from it-like, and such-Anglo-Saxon, swilk-Gothic, swa-leiks-from so-like.

The demonstrative prefix is in general t, d, or th, and thus are formed, in German, der, das, from er, es; in English, that from it:— but sometimes the demonstrative pronoun has an initials, as in

the

the Gothic,-sa, so, thata, Anglo-Saxon, se, seo, that. The proper use of the demonstrative pronoun that seems to be to single out some object in the presence of the speaker which he wishes to make the subject of discourse, and to distinguish it from others of a like nature. The simplest mode of doing this would be by actually pointing it out, at the same time making use of some expression in order to call the hearer's attention to the action of pointing; and this expression, whatever it might be, would correspond in construction to the demonstrative pronoun. It will be evident that such is the office of this pronoun from considering the case where there is no accompanying action. If there are several books on a table, and I desire some one to bring me that book,' without making use of any action, it will be impossible for him to know which book I want, or to attribute any meaning to the word that. It is thus by the action alone that we are directed to the object intended; and the use of the demonstrative can only be to call the hearer's attention to that action. Now we have seen that that is composed of a particle t or th prefixed to the personal pronoun it; and the latter word, meaning simply thing, the demand of the hearer's attention must be made by the prefix t, which would thus be the vestige of some such imperative as look or see. The latter would agree very well with the Anglo-Saxon se; and it is probable that the initial t in the, that, has the same origin, since s and t are interchanged in innumerable instances, as in das and that, Tooaga and Terraga, ou and tu, &c. &c.

In his account of the demonstratives, Grimm says that this prefix in Latin exists only in tam, tantus, tot, and talis: he seems to have overlooked is-te, a peculiarly instructive word; as we see here that the force of the demonstrative particle te resides in its very form, and that it is not necessary to its effect that it should be placed at the beginning of a word. From being used to point out some object actually present, it was an easy step to employ the demonstrative pronoun for the purpose of defining and individualizing any object to which, though absent, it was wished to call the attention of the person spoken to. In this case it is obvious that, as there can be no action by which the individual intended can be designated, some distinguishing circumstances must be mentioned by word of mouth, and then the office of the pronoun (which will now assume the form of the definite article) will be to call the hearer's attention to these circumstances, the mention of which corresponds to the actual pointing out of a present object. Thus, if I say, 'Bring me that book out of the library,' I must add some circumstances to distinguish the book I want from the rest. I may thus say, 'Bring me that book in scarlet binding,' or ' Bring me the book that is lying on its side,'

where

where the mention of the distinguishing circumstances corresponds to the action of pointing; and the use of the words that and the is merely to draw the hearer's attention to these circumstances. It is true that in English we have different words-that and thefor what we have called the original use of the demonstrative pronoun, and for the definite article, but it is clear that this is a modern refinement, as the same word was used by the Anglo-Saxons, and still is by the Germans in both senses, and our demonstrative that is merely the neuter of the Anglo-Saxon article se. In many cases in English, and much more so in German, this pronoun has usurped the place of the relative. Thus, instead of

'Light is a body which moves rapidly,'

we might with equal propriety say,

'Light is a body that moves rapidly.'

The first of these examples is, as we have seen, an elliptic expres

sion for

Light is a body, the particular body spoken of moves rapidly.' But we render this more emphatically—

'Light is a body, and that [body]" (pointing it out, as it were, with the finger from among all other bodies)" moves rapidly;'

or elliptically

Light is a body that moves rapidly.'

And thus it is that the demonstrative is so frequently used in a relative sense.

The adverbs then and there answer so exactly to their correlatives when, where, that it would be useless to say anything of them here. The particle than, however, has in English no corresponding relative, and it appears at first sight to partake so strongly of a comparative sense, that we have some difficulty in believing it to be merely a case of the demonstrative pronoun the. Grimm

says

The accusative nature of the Gothic particle than is clearly shown by the Latin tum (as eum, illum) and tunc (for tum-c-Gothic thanuh). The particle than is, indeed, not confounded with the accusative of the pronoun thana (eum), although even here the a is omitted in one passage (Mark. 15, 44). On the other hand, the a has been preserved to the adverb in composition in thana-mais, thana-seiths, and I think it is clear that than arises out of thana. The Gothic than signifies TOTE, then-sometimes it has the abstract sense of ovv, igitur, of yap, still more commonly of de, vero. The suffix uh (corresponding to the Latin c) does not alter these meanings: thanuh and than, as tum and tunc, being of like import, except that thanuh generally occurs at the beginning, than in the middle, of a sentence. In old high German the corresponding particle danne, besides these meanings, also expresses quam after a comparative. In Anglo-Saxon the accu

sative is thone, the particle thon or thonne: thon corresponding to the Gothic than, thon-ne to than-uh, thon and thonne signify tum, tunc,seldomer dum, quando,—but thonne alone is used to express quam after a comparative.' (III. 165.)

Thus, in any comparison, such as Richard is wiser than John,' we are compelled, by the derivation of the word than, to attribute to it the meaning of that only; and the sentence might as well be expressed (as it is in French) Richard is wiser that John.' The parts of grammar are so closely interwoven together, that it is impossible to explain the force of the word than in this situation, without some previous examination of the nature of the comparative degree in adjectives. In this pursuit we may derive a useful hint from the Chinese, a language so rude that it has not arrived at the use of a single inflection; but wherein all the modifications which are in other languages effected by this means, take place by means of separate words so that even the plural number in nouns is formed by adding the word men to the singular. This is probably the state in which all languages originally were, and all inflections must have taken their rise from some such auxiliary words. Now as the fundamental idea represented by the original form of a word is always present in each of its inflected forms, it is clear that what we have called the modification caused by any inflexion must in reality be the addition of a new idea, the exact nature of which it is often extremely difficult to discover; and in this investigation it will obviously be of the greatest service to us, if we can find a language where a corresponding effect is produced by an auxiliary word, as we shall then have two distinct ideas into which the modified idea represented by the inflected word may be divided. Now in Chinese, we are told that the character ko or kwo, to pass beyond, is generally used to express the comparative degree in adjectives. Thus, Hou yoong kwo gno is, in that language, He is more vehement than I,' literally, He is vehement beyond me.' On comparing this mode of expressing the comparative degree with our own, we shall find that the syllable er, ter, or ther, by the addition of which to the simple adjective we form our comparative, conveys the idea of excess, and thus the word wiser represents two ideas, that of wisdom, and that of excess. Now, every comparison of two things necessarily includes the supposition, that they both possess in some degree the quality with respect to which they are compared, and a judgment is then pronounced, that the degree in which one of the objects compared possesses it, exceeds that of the other. Thus, in the example above given-Richard is wiser than John'—we may supply the assertion that John is wise to a

1

2 3

2

3

certain