affix the particle not to can for the purpose of expressing, He cannot (do any thing), or he can do nothing ONLY fail. Both forms of expression may be regarded as idiomatic and elliptical, and as expressing a sense which could scarcely be ascertained by a mere reference to the ordinary laws of language. We must in addition have recourse to the history of the language, if we would trace such usages successfully. In such cases, the whole expression has come, by conventional usage, to bear a meaning which is not to be discovered by ascertaining the proper import of the several words and their construction. This seems to us the proper description of idiomatic expressions. In forming all such, the laws of language-the fixed principles which have generally guided its development-have been held in abeyance, and a conventional sanction has been given to the products of accident, or of the caprice of fashion.

We see no reason, because of such examples as that now considered, for refusing to acquiesce in the account of this singular use of the word BUT given by Mr. Tooke, and sanctioned by Dr. Webster and other lexicographers and grammarians. We cannot give the same ready acquiescence, as will be seen presently, to some other opinions advanced by Mr. Tooke in reference to this word.

151. The word BUT performs such important and various and apparently dissimilar functions in our language, that it has given occasion to much speculation among grammarians and philologists. For these reasons, we subjoin a few additional observations, in which we shall attempt to trace the steps by which it has made the transition from the earliest use, with which we are acquainted, to that which is apparently its latest, and, certainly, its most common use at the present time. We are the more induced to do this, because the accounts commonly given of the origin of the adversative and exceptive uses of BUT do not appear to us completely satisfactory. We cannot acquiesce in some of the rash conclusions which Mr. H. Tooke defends in his usual dictatorial manner. We shall subject his reasonings to a brief examination, reject what appears to us inconclusive, and endeavor to trace the probable history of the transition from the exceptive to the adversative use of this very important word.

The word but was employed in the earliest times in a sense distinct from any in which it is now employed. It seems to have been nearly equivalent to without, and to have had a just claim to be considered a preposition.

Mr. Tooke, in his attempt to establish a distinction between but and bot, has furnished a host of examples of this apparently original use of but from Gawin Douglas's Translation of the Eneid. We give a few of these examples in the briefest manner consistent with our purpose, and refer the inquisitive reader to the Diversions of Purley (article on the word but), for the passages in connection with these examples.

"Thare is gret substance ordanit the BUT dout."

In modern English, There is great substance (wealth, fortune) ordained (decreed) to thee WITHOUT doubt.

"And als mony nychtes BUT Sterneys leme."

And as many nights WITHOUT star-light.

"Before Eneas feite stude, BUT delay."

= Before Æneas' feet stood wrrnout delay.”

"Bot of the bargane maid end, BUT delay."

We have followed the pointing of Mr. Taylor's edition of the Diversions of Purley. We see no reason, however, for using the comma in the last two examples, if it is unnecessary in the first two.

In all these passages BUT is employed, like a preposition, before single nouns. It connects a word-complement, not a proposition-complement. Mr. Tooke draws no distinction between this use of but and its still current use, to express exception. No doubt, the two uses are allied, and we believe the present exceptive use has originated from the more ancient use exhibited above. Still there is a marked transition from the one to the other. In not one of these passages could except or save be substituted for BUT, as they can generally, if not universally for but exceptive. Compare with these examples, John has learnt all his lessons BUT one. "All but the wakeful night

ingale." Here but is manifestly equivalent to except, save.

Examples of the use of BUT similar to those above from Gawin Douglas are, we believe, to be found chiefly in Scotch authors. This use is rarely found in the writers of South Britain, of even the earliest date.

Mr. Tooke has attempted to prove that BUT adversative, employed to connect a proposition in some respect opposed to the tenor of the preceding discourse, is a word entirely distinct from BUT employed for the other purposes above mentioned; and his conclusions in reference to this matter have been followed by many since his time, apparently without much examination. This adversative but is, according to Mr. Tooke's view, the imperative of an Anglo-Saxon verb botan to boot, and was spelled bot; whereas but is the imperative of be-utan, to be out. In reference to Mr. Locke, he remarks, "It was the corrupt use of this one word (BUT) in modern English, for two words (BOT and BUT) originally (in the Anglo-Saxon) very different in signification, though (by repeated abbreviation and corruption) approaching in sound, which chiefly misled him." By this passage we might be led to conclude that bot, distinct from but, is found employed as a conjunction in the Anglo-Saxon language. The word bot or bote meaning boot, advantage, remedy, is indeed found in Anglo-Saxon employed as a noun. But we doubt whether this word was ever used alone in Anglo-Saxon or old English, as a connective. (To bote, is sometimes used to mean moreover.) As to the verb botan, and its imperative bot or bote, we can find no reason to believe that they ever existed, except in the imaginations of Mr. Tooke and some of his followers.

We suspect that bot in Gawin Douglas is only a variation in spelling, the orthography of those times being, as Mr. Tooke himself admits, very unsettled. Mr. Richardson in his dictionary has given examples of bot and bote from Robert of Gloucester and others; but in these examples there is no distinction maintained between but and bot or bote. Bot and bote are used on several occasions, when the sense is plainly butan, except. As, for example,

"Ne that no man ys wurthe to be ycluped (called) Kyng.
Bote the heye king of heuene," &c.

"That in all the lond suld be no king bot he."

except, and not for the

Bote and bot are plainly here employed for but adversative but. Such examples show that the authors, to whom Mr. Tooke and Mr. Richardson appeal, are not to be relied on to establish the different origin of the adversative but, and but without and except, since they spell the word in all these senses indifferently bot or bote or but. Mr. Richardson himself, as well as the authors whom he quotes, seems to confound these meanings of but, which he meant to separate. He has arranged one, if not two, examples under but or bot, which should have been, on his principles, placed under but. We refer to the citations from Holland, Plinie, &c., and from Feltham. The last is perhaps doubtful. See Richardson's Dict.

Art. But.

Examples of butan or buton used in the adversative sense of our modern but are not, we suspect, to be found in Anglo-Saxon. The connection which we express by but adversative seems to have been expressed invariably in that language by ac; never by bot, as an incautious reader of the Diversions of Purley might be led to suppose. No trace of ac thus used (except it is in eke) remains in our language. Its place seems to have been supplied not by bot from botan-the invention of Mr. Tooke-but by an insensible extension of the exceptive but. The transition from the exceptive use of but to but adversative, in cases where it has the greatest force, does not appear to us so violent as Mr. Tooke represents it. Let us take for example that passage in the 115th Psalm; "They have mouths, but they speak not, eyes have they, but they see not," &c. The transition from the exceptive to the use here found does not seem to us much wider than from the use equivalent to the preposition without to the exceptive sense. And yet Mr. Tooke thinks these so much alike that he does not recognise them as distinct. By but exceptive we except a complementary word, by but adversative a proposition. The first, in other words, indicates an exception from a proposition, the latter an exception from the discourse-from the train of thought. Mr. Tooke seems to us to have been led, by his attempt to support a foregone conclusion, to rob but adversative of an essential part of its significance. It fits his purpose to make it mean only add, and in this way he has entirely destroyed all distinction between this word and AND.

It may be added that Mr. Tooke admits that ancient authors have not consistently observed the distinction between bot and but. He confesses that no trace of it can be found in Chaucer, and we can find none in Wiclif's New Testament, printed by Bagster from a good manuscript.

We certainly see no necessity for the high-handed course pursued by Mr. Tooke in order to account for the distinction between but adversative and but exceptive. But we do not presume to pronounce a positive opinion. We leave the question to be settled by those who possess sufficient acquaintance with Anglo-Saxon, and old English literature, in which we do not pretend to be versed. We regard Mr. Tooke's rash and confident ipse dixit as of no weight in the case before us, and consider the question open for discussion.

In the mean time we have traced the transition of but from one meaning to another, so far as we have been able to do it by the helps accessible to us. To sum up the results we have,

1st. BUT- WITHOUT. This is likely the primary use of this word, and likely enough, as H. Tooke and others allege, both the Anglo-Saxon butan or buton and this are the imperative be-utan of beon-utan. It is now found almost exclusively in the older Scotch writers, and is, perhaps, the same word still used adverbially in the colloquial Scotch in the expression gang go out.

BUT=

2d. BUT EXCEPT. This meaning, like the last, is usually traced to the original sense of the compound be-utan, be out — except. Both these meanings are expressed in the Anglo-Saxon by butan or buton.

3d. BUT ONLY. This is the same as the last when the negative is sup plied. So far we accept the account given by Mr. Tooke & Co.

4th. BUT Adversative. This indicates, as we have already shown, the addition or connection of something variant from what precedes; or perhaps we may venture to say an exception in reference to the train of the discourse, whereas but in the second sense indicates an exception in reference to the proposition in which it is employed.

There remains a point yet to be settled. With what class of words shall we arrange BUT Exceptive? When employed in the first-mentioned manner as equivalent to without, the Anglo-Saxon grammarians call it a preposition. and class it with those which take after them a dative case; with this use we have nothing to do in this place, as it is long since entirely obselete.

Bur exceptive has also been recognised by many of our late grammarians, and by some lexicographers as a preposition. But if but exceptive is a preposition, it should, according to the definition which these same grammarians give of a preposition, take the noun which it connects always in the acccusative This we think it never does, according to the most reputable usage of the language, except when the accusative which follows is plainly the objec tive modification of a verb suppressed. For example, we consider, There is nobody here BUT I, the true construction, and not, There is nobody here BUT

case.

ME.

When this expression is fully developed, it would be, There is nobody here BUT I am here except I am here. When the verb suppressed is an active verb, and the noun after but would, if the verb were supplied, stand to it in the relation of objective modification, we place this noun in the accusative thus, I saw nobody BUT HIM = I saw nobody but, or, except I saw him. That is, I saw nobody BE OUT I saw him. In this form I saw him is subject to the conjunctive verb BE OUT.

We believe that examples may be found in reputable authors of the use of an accusative dependent on but alone; but we think they are not to be imitated. We subjoin a few examples of what we consider the correct usage of the language in this matter, as well as the general usage of educated speakers and writers. The intelligent reader will readily perceive that the oider these examples are the more weight they have in determining the question before us. If old, they prove that the present usage is not a modern refinement of the educated classes, or based on grammatical theories, but the effect of spontaneous development. We adduce first a line already quoted from one of our oldest English authors.

"That in alle the lond suld be no king but HE."

"There is none other but he." Mark 12: 32. He hath not grieved me but in part." "Neither was I tanght it but by the revelation of Jesus Christ." In all these examples, whether we say that but connects a complementary proposition, or that it only connects a complementary word, it is obvious that it has not the characteristic of a preposition. In the first two examples it is followed by a nominative case, in the last two by a preposition.

It is not perhaps uncommon to hear persons use me after but where me cannot be the objective modification of any other word. Perhaps the same persons might be found to say, there is nobody in the house but he, and yet say, there is nobody here but me. This may be attributable to the fact to be noticed in another place (See § 155: 12), that in the language of the uneducated me is used as nominative in all cases where it is not manifestly the immediate subject-noun to a verb. Thus, the answer of the uneducated classes to the question, Who is there? is universally ME. In the same man. ner the accusative (or rather in this case we should say the dative) is still used by the uneducated after comparatives and than, as He is taller THAN ME We may place these expressions on a level with, There is nobody here BUT ME, as regards purity of language. The dative after than has perhaps stronger claims to be recognised as English than the accusative after but; and consequently than has as great if not greater claims than but to be considered sometimes a preposition. We may cite Milton as authority for the use of a dative or accusative after than.

"Belial came last, THAN WHOм a spirit more lewd

Fell not from Heaven," &c.

The present usage of correct writers and speakers, we believe, is to avoid