I am persuaded, however, that it arises in part from our surprise at the poet's habits of association, which enable him to convey his thoughts with ease and beauty, notwithstanding the narrow limits within which his choice of expression is confined. One proof of this is, that if there appear any mark of constraint, either in the ideas or in the expression, our pleasure is proportionally diminished. The thoughts must seem to suggest each other, and the rhymes to be only an accidental circumstance. The same remark may be made on the measure of When in its greatest perfection, it does not appear to be the result of labour, but to be dictated by nature, or prompted by inspiration. In Pope's best verses, the idea is expressed with as little inversion of style, and with as much conciseness, precision, and propriety, as the author could have attained, had he been writing prose; without any apparent exertion on his part, the words seem spontaneously to arrange themselves in the most musical numbers.

the verse.

"While still a child, nor yet a fool to fame,

I lisp'd in numbers, for the numbers came."

This facility of versification, it is true, may be, and probably is, in most cases, only apparent; and it is reasonable to think, that in the most perfect poetical productions, not only the choice of words, but the choice of ideas, is influenced by the rhymes. In a prose composition, the author holds on in a direct course, according to the plan he has previously formed; but in a poem, the rhymes which occur to him are perpetually diverting him to the right hand or to the left, by suggesting ideas which do not naturally rise out of his subject. This, I presume, is Butler's meaning in the following couplet :—

Rhymes the rudder are of verses

With which, like ships, they steer their courses."

But although this may be the case in fact, the poet must employ all his art to conceal it; insomuch, that if he finds himself under a necessity to introduce, on account of the rhymes, a superfluous idea, or an awkward expression, he must place it in the first line of the couplet, and not in the second; for the reader, naturally presuming that the lines were composed in

the order in which the author arranges them, is more apt to suspect the second line to be accommodated to the first, than the first to the second. And this slight artifice is, in general, sufficient to impose on that degree of attention with which poetry is read. Who can doubt that, in the following lines, Pope wrote the first for the sake of the second?

A wit's a feather, and a chief a rod;

An honest man's the noblest work of God."

Were the first of these lines, or a line equally unmeaning, placed last, the couplet would have appeared execrable to a person of the most moderate taste.1

It affords a strong confirmation of the foregoing observations, that the Poets of some nations have delighted in the practice of alliteration, as well as of rhyme, and have even considered it as an essential circumstance in versification. Dr. Beattie observes, that "some ancient English poems are more distinguished by alliteration, than by any other poetical contrivance. In the works of Langland, even when no regard is had to rhyme, and but little to a rude sort of anapæstic measure, it seems to have been a rule, that three words, at least, of each line should begin with the same letter." A late author informs us, that, in the Icelandic poetry, alliteration is considered as a circumstance no less essential than rhyme.

He mentions also several

one of his epistles goes far to justify
the foregoing sarcasm.

Antoine, gouverneur de mon jardin d'Autueil,
Qui dirige chez moi l'if et le chèvrefeuille.]

"The Icelandic poetry requires two things, namely, words with the same initial letters, and words of the same sound. It was divided into stanzas, each of which consisted of four couplets; and each of these couplets was again composed of two hemistichs, of which every one contained six syllables; and it was not allowed to augment this num ber, except in cases of the greatest necessity."-See Van Troil's Letters on Iceland, p. 208.

other restraints, which must add wonderfully to the difficulty of versification, and which appear to us to be perfectly arbitrary and capricious. If that really be the case, the whole pleasure of the reader or hearer arises from his surprise at the facility of the Poet's composition under these complicated restraints, that is, from his surprise at the command which the Poet has acquired over his thoughts and expressions. In our rhyme, I acknowledge that the coincidence of sound is agreeable in itself, and only affirm, that the pleasure which the ear receives from it is heightened by the other consideration.

3. OF POETICAL FANCY.

There is another habit of association which, in some men, is very remarkable-that which is the foundation of Poetical Fancy a talent which agrees with Wit in some circumstances, but which differs from it essentially in others.

The pleasure we receive from wit, agrees in one particular with the pleasure which arises from poetical allusions,—that in both cases we are pleased with contemplating an analogy between two different subjects. But they differ in this, that the man of wit has no other aim than to combine analogous ideas; whereas no allusion can, with propriety, have a place in serious poetry, unless it either illustrate or adorn the principal subject. If it has both these recommendations, the allusion is perfect. If it has neither, as is often the case with the allusions of Cowley and of Young, the Fancy of the Poet degenerates into wit.

If these observations be well-founded, they suggest a rule with respect to poetical allusions, which has not always been sufficiently attended to. It frequently happens that two subjects bear an analogy to each other in more respects than one; and where such can be found, they undoubtedly furnish the most favourable of all occasions for the display of wit. But, in serious poetry, I am inclined to think, that however striking

I speak here of pure and unmixed wit-and not of wit blended, as it is

most commonly, with some degree of humour.

these analogies may be, and although each of them might with propriety be made the foundation of a separate allusion, it is improper, in the course of the same allusion, to include more than one of them, as, by doing so, an author discovers an affectation of wit, or a desire of tracing analogies, instead of illustrating or adorning the subject of his composition.1

I formerly defined Fancy to be a power of associating ideas according to relations of resemblance and analogy. This definition will probably be thought too general, and to approach too near to that given of wit. In order to discover the necessary limitations, we shall consider what the circumstances are which please us in poetical allusions. As these allusions are suggested by Fancy, and are the most striking instances in which it displays itself, the received rules of critics with respect to them, may throw some light on the mental power which gives them birth.

1. An allusion pleases, by illustrating a subject comparatively obscure. Hence, I apprehend, it will be found, that allusions from the intellectual world to the material, are more pleasing than from the material world to the intellectual. Mason, in his Ode to Memory, compares the influence of that faculty over our ideas, to the authority of a general over his troops :

[In the following stanza of Shenstone, for example,

"How pale was then his true-love's cheek When Jemmy's sentence reach'd her ear! For never yet did Alpine snows

So pale, or yet so chill appear;" the double allusion unquestionably borders on conceit. The same double allusion occurs in the translation of Mallet's "William and Margaret," by Vincent Bourne,

"Candidior nive, frigidiorque manus." How inferior in pathetic simplicity to the original,

"And clay-cold was the lily hand," &c. That Shenstone himself considered

these double allusions as more allied to wit than to the language of serious passion, appears from the style of poetry ascribed to Paridel in the Pastoral Ballad.

""Tis his with mock passion to glow;

"Tis his in smooth tales to unfold, How her face is as bright as the snow,

And her bosom, be sure, is as cold."

Mr. Addison's opinion is of still higher value. "When a poet tells us, the bosom of his mistress is as white as snow, there is no wit in the comparison; but when he adds, with a sigh, that it is as cold too, it then grows into wit." -Spectator, No. 62.]

-"thou, whose sway

The throng'd ideal hosts obey;

Who bidst their ranks now vanish, now appear;

Flame in the van, or darken in the rear."

Would the allusion have been equally pleasing, from a general marshalling his soldiers, to Memory and the succession of ideas?

The effect of a literal and spiritless translation of a work of genius, has been compared [by Cervantes] to that of the figures which we see, when we look at the wrong side of a beautiful piece of tapestry.1 The allusion is ingenious and happy; but the pleasure which we receive from it arises, not merely from the analogy which it presents to us, but from the illustration which it affords of the author's idea. No one, surely, in speaking of a piece of tapestry, would think of comparing the difference between its sides, to that between an original composition and a literal translation!

Cicero, and after him Mr. Locke, in illustrating the difficulty of attending to the subject of our consciousness, have compared the Mind to the Eye, which sees every object around it, but is invisible to itself. To have compared the Eye, in this respect, to the Mind, would have been absurd.

Mr. Pope's comparison of the progress of youthful curiosity, in the pursuits of science, to that of a traveller among the Alps, has been much and justly admired. How would the beauty of the allusion have been diminished, if the Alps had furnished the original subject and not the illustration!

But although this rule holds in general, I acknowledge that instances may be produced from our most celebrated poetical performances, of allusions from material objects, both to the intellectual and the moral worlds. These, however, are comparatively few in number, and are not to be found in descrip

["For all that, I cannot but be of opinion, that translating out of one language into another, unless it be from those queens of the languages, Greek and Latin, is like setting to view the wrong side of a piece of tapestry, where,

though the figures are seen, they are full of ends and threads, which obscure them, and are not seen with the smoothness and evenness of the right side."Don Quixote, chap. xii. Jarvis's Translation.]