the accidents of life, by heaping up treasures, which nothing can protect us against, but the good providence of our heavenly Father.' Which, always refers grammatically to the immediately preceding substantive, which is here treasures;' and this would make nonsense of the whole period. Every one feels this impropriety. The sentence ought to have stood thus: It is folly to pretend, by heaping up treasures, to arm ourselves against the accidents of life, which nothing can protect us against but the good providence of our heavenly Father.'

Of the like nature is the following inaccuracy of Dean Swift's. He is recommending to young clergymen, to write their sermons fully and distinctly. Many,' says he, act so directly contrary to this method, that, from a habit of saving time and paper, which they acquired at the university, they write in so diminutive a manner, that they can hardly read what they have written.' He certainly does not mean, that they had acquired time and paper at the university, but that they had acquired this habit there; and therefore his words ought to have run thus: From a habit, which they have acquired at the university, of saving time and paper, they write in so diminutive a manner.' In another passage, the same author has left his meaning altogether uncertain, by misplacing a relative. It is in the conclusion of his letter to a member of parliament, concerning the sacramental test: Thus I have fairly given you, Sir, my own opinion, as well as that of a great majority of both houses here, relating to this weighty affair; upon which I am confident you may securely reckon.' Now I ask, what it is he would have his correspondent to reckon upon, securely? The natural construction leads to these words, 'this weighty affair.' But, as it would be difficult to make any sense of this, it is more probable he meant that the majority of both houses might be securely reckoned upon; though certainly this meaning, as the words are arranged, is obscurely expressed. The sentence would be amended by arranging it thus: Thus, Sir, I have given you my own opinion, relating to this weighty affair, as well as that of a great majority of both houses here; upon which I am confident you may securely reckon.'

Several other instances might be given; but I reckon those which I have produced sufficient to make the rule understood; that, in the construction of sentences, one of the first things to be attended to, is the marshalling of the words in such order as shall most clearly mark the relation of the several parts of the sentence to one another; particularly, that adverbs shall always be made to adhere closely to the words which they are intended to qualify; that, where a circumstance is thrown in, it shall never hang loose in the midst of a period, but be determined by its place to one or other member of it; and that every relative word which is used, shall instantly present its antecedent to the mind of the reader, without the least obscurity. I have mentioned these three cases, because I think they are the most frequent occasions of ambiguity creeping into sentences.

With regard to relatives, I must further observe, that obscurity often arises from the too frequent repetition of them, particularly of

the pronouns who, and they, and them, and theirs, when we have occasion to refer to different persons; as, in the following sentence of Archbishop Tillotson; (vol. 1. serm. 42.) 'Men look with an evil eye upon the good that is in others; and think that their reputation obscures them, and their commendable qualities stand in their light; and therefore they do what they can to cast a cloud over them, that the bright shining of their virtues may not obscure them.' This is altogether careless writing. It renders style often obscure, always embarrassed and inelegant. When we find these personal pronouns erowding too fast upon us, we have often no method left, but to throw the whole sentence into some other form, which may avoid those frequent references to persons who have before been mentioned.

All languages are liable to ambiguities. Quintilian gives us some instances in the Latin, arising from faulty arrangement. A man, he tells us, ordered by his will, to have erected for him, after his death, 'Statuam auream hastam tenentem;' upon which arose a dispute at law, whether the whole statue, or the spear only, was to be of gold? The same author observes, very properly, that a sentence is always faulty, when the collocation of the words is ambiguous, though the sense can be gathered. If any one should say, Chremetem audivi percussisse Demeam,' this is ambiguous, both in sense and structure, whether Chremes or Demea gave the blow. But if this expression were used, 'Se vidisse hominem librum scribentem,' although the meaning be clear, yet Quintilian insists that the arrangement is wrong. Nam,' says he, 'etiamsi librum ab homine scribi pateat, non certè hominem a libro, malè tamen composuerat, feceratque ambiguum quantum in ipso fuit.' Indeed, to have the relation of every word and member of a sentence marked in the most proper and distinct manner, gives not clearness only, but grace and beauty to a sentence, making the mind pass smoothly and agreeably along all the parts of it.

I proceed now to the second quality of a well-arranged sentence, which I termed its unity. This is a capital property. In every composition, of whatever kind, some degree of unity is required, in order to render it beautiful. There must be always some connecting principle among the parts. Some one object must reign and be predominant. This, as I shall hereafter show, holds in history, in epic and dramatic poetry, and in all orations. But most of all, in a single sentence, is required the strictest unity. For the very nature of a sentence implies one proposition to be expressed. It may consist of parts, indeed; but these parts must be so closely bound together, as to make the impression upon the mind, of one object, not of many. Now, in order to preserve this unity of a sentence, the following rules must be observed:

In the first place, during the course of the sentence, the scene should be changed as little as possible. We should not be hurried by sudden transitions from person to person, nor from subject to subject. There is commonly, in every sentence, some person or thing, which is the governing word. This should be continued so, if possible, from the beginning to the end of it. Should I express

myself_thus: 'After we came to anchor, they put me on shore, where I was welcomed by all my friends, who received me with the greatest kindness.' In this sentence, though the objects contained in it have a sufficient connexion with each other, yet, by this manner of representing them, by shifting so often both the place and the person, we, and they, and I, and who, they appear in such a disunited view, that the sense of connexion is almost lost. The sentence is restored to its proper unity, by turning it after the following manner: Having come to an anchor, I was put on shore, where I was welcomed by all my friends, and received with the greatest kindness.' Writers who transgress this rule, for the most part transgress, at the same time,

A second rule; never to crowd into one sentence, things which have so little connexion, that they could bear to be divided into two or three sentences. The violation of this rule never fails to hurt and displease a reader. Its effect, indeed, is so bad, that of the two, it is the safer extreme, to err rather by too many short sentences, than by one that is overloaded and embarrassed. Examples abound in authors. I shall produce some to justify what I now say. Archbishop Tillotson, says an author of the History of England, died in this year. He was exceedingly beloved both by king William and queen Mary, who nominated Dr. Tennison, Bishop of Lincoln, to succeed him.' Who would expect the latter part of this sentence to follow, in consequence of the former? He was exceedingly beloved by both king and queen,' is the proposition of the sentence: we look for some proof of this, or at least something related to it to follow; when we are on a sudden carried off to a new proposition, who nominated Dr. Tennison to succeed him.' The following is from Middleton's Life of Cicero: In this uneasy state, both of his public and private life, Cicero was oppressed by a new and cruel affliction, the death of his beloved daughter Tullia; which happened soon after her divorce from Dolabella; whose manners and humours were entirely disagreeable to her.' The principal object in this sentence is, the death of Tullia, which was the cause of her father's affliction; the date of it, as happening soon after her divorce from Dolabella, may enter into the sentence with propriety; but the subjunction of Dolabella's character is foreign to the main object; and breaks the unity and compactness of the sentence totally, by setting a new picture before the reader. The following sentence, from a translation of Plutarch, is still worse: Their march,' says the author, speaking of the Greeks under Alexander, 'their march was through an uncultivated country, whose savage inhabitants fared hardly, having no other riches than a breed of lean sheep, whose flesh was rank and unsavoury, by reason of their continual feeding upon sea-fish.' Here the scene is changed upon us again and again. The march of the Greeks, the description of the inhabitants through whose country they travelled, the account of their sheep, and the cause of their sheep being ill-tasted food, form a jumble of objects, slightly related to each other, which the reader cannot, without much difficulty, comprehend under one view.

These examples have been taken from sentences of no great length, yet over-crowded. Authors who deal in long sentences, are very apt to be faulty in this article. One need only open Lord Clarendon's history, to find examples every where. The long, involv ed, and intricate sentences of that author, are the greatest blemish of his composition; though, in other respects, as a historian, he has considerable merit. In later, and more correct writers than Lord Clarendon, we find a period sometimes running out so far, and comprehending so many particulars, as to be more properly a discourse than a sentence. Take, for an instance, the following, from Sir William Temple, in his Essay upon Poetry: "The usual acceptation takes profit and pleasure for two different things; and not only calls the followers or votaries of them by the several names of busy and idle men; but distinguishes the faculties of the mind, that are conversant about them, calling the operations of the first, wisdom; and of the other, wit; which is a Saxon word, used to express what the Spaniards and Italians call ingenio, and the French, esprit, both from the Latin; though I think wit more particularly signifies that of poetry, as may occur in remarks on the Runic language.' When one arrives at the end of such a puzzled sentence, he is surprised to find himself got to so great a distance from the object with which he at first set out.

Lord Shaftesbury, often betrayed into faults by his love of magnificence, shall afford us the next example. It is in his rhapsody where he is describing the cold regions: At length,' says he, 'the sun approaching, melts the snow, sets longing men at liberty, and affords them means and time to make provision against the next return of cold;' This first sentence is correct enough; but he goes on: 'It breaks the icy fetters of the main, where vast sea-monsters pierce through floating islands, with arms which can withstand the crystal rock; whilst others, who of themselves seem great as islands, are by their bulk alone armed against all but man, whose superiority over creatures of such stupendous size and force, should make him mindful of his privilege of reason, and force him humbly to adore the great composer of these wondrous frames, and the author of his own superior wisdom.' Nothing can be more unhappy or embarrassed than this sentence; the worse, too, as it is intended to be descriptive, where every thing should be clear. It forms no distinct image whatever. The it, at the beginning, is ambiguous, whether it mean the sun or the cold. The object is changed three times in the sentence; beginning with the sun, which breaks the icy fetters of the main; then the sea-monsters become the principal personages; and lastly, by a very unexpected transition, man is brought into view, and receives a long and serious admonition, before the sentence closes. I do not at present insist on the impropriety of such expressions as, God's being the composer of frames; and the sea-monsters having arms that withstand rocks. Shaftesbury's strength lay in reasoning and sentiment, more than in description; however much his descriptions have been sometimes admired.

I shall only give one instance more on this head, from Dean Swift;

in his proposal, too, for correcting the English language: where, in place of a sentence, he has given a loose dissertation upon several subjects. Speaking of the progress of our language, after the time of Cromwell: To this succeeded,' says he, that licentiousness which entered with the restoration, and from infecting our religion. and morals, fell to corrupt our language; which last was not likely to be much improved by those, who at that time made up the court of king Charles the Second; either such as had followed him in his banishment, or who had been altogether conversant in the dialect of these fanatic times; or young men who had been educated in the same country; so that the court, which used to be the standard of correctness and propriety of speech, was then, and I think has ever since continued, the worst school in England for that accomplishment; and so will remain, till better care be taken in the education of our nobility, that they may set out into the world with some foundation of literature, in order to qualify them for patterns of politeness.'How many different facts, reasonings, and observations, are here presented to the mind at once! and yet so linked together by the author, that they all make parts of a sentence, which admits of no greater division in pointing, than a semicolon between any of its members? Having mentioned pointing, I shall here take notice, that it is in vain to propose, by arbitrary punctuation, to amend the defects of a sentence, to correct its ambiguity, or to prevent its confusion. For commas, colons, and points, do not make the proper divisions of thought; but only serve to mark those which arise from the tenour of the author's expression; and, therefore, they are proper or not, just according as they correspond to the natural division of the sense. When they are inserted in wrong places, they deserve, and will meet with, no regard.

I proceed to a third rule, for preserving the unity of sentences, which is, to keep clear of all parentheses in the middle of them. On some occasions, these may have a spirited appearance; as prompted by a certain vivacity of thought, which can glance happily aside, as it is going along. But, for the most part, their effect is extremely bad; being a sort of wheels within wheels; sentences in the midst of sentences; the perplexed method of disposing of some thought, which a writer wants art to introduce in its proper place. It were needless to give many instances, as they occur so often among incorrect writers. I shall produce one from Lord Bolingbroke; the rapidity of whose genius, and manner of writing, betrays him frequently into inaccuracies of this sort. It is in the introduction to his idea of a patriot king, where he writes thus: 'It seems to me, that, in order to maintain the system of the world, at a certain point, far below that of ideal perfection, (for we are made capable of conceiving what we are incapable of attaining) but, however, sufficient, upon the whole, to constitute a state easy and happy, or, at the worst, tolerable; I say, it seems to me, that the Author of Nature has thought fit to mingle, from time to time, among the societies of men, a few, and but a few, of those on whom he is graciously pleased to