exercised in the different subjects of their investigation, but must have reduced his speculations in them under such general heads, as will give his thoughts the consistence and utility, which arise from system. He cannot hope, without being somewhat of a good historian, to succeed in those higher walks of his art, which take their subject from the oral or written annals of a nation. Without much of the skill and observation of a naturalist, his descriptions of rural scenery, and his delineation of life and manners, must be cold and uninteresting. That knowledge of the human heart and character, of the calm tenour of sentiment, and the warm ebullition of passion, which he is so frequently called upon to display, he must derive from the same study whence the moral philosopher constructs his system of ethicks. Over language in all the varieties of sense, structure, embellishment, and harmony, he must exercise the skill of a grammarian and a rhetorician. And he must complete his education in this circle of sciences, by acquiring a perfect insight into those critical rules by which his art is to be tried on the touchstone of excellence.

But though the poet is thus brought within the fence of science, he is not confined to the narrow limits of its circumference. By a certain felicity of boldness, which has ever been the undisputed right of his art, he may break down that pale which would set bounds to his prerogative. To the language of history he is not always obliged to pay a rigid attention; he may often give to past events, a turn which is more suitable to the elevation of his ideas; and may represent things, not as they happened, but as he conceives they might have happened. In his delineation of natural scenery, and pictures of human life and action, descriptions and cha -racters may rise from his creation, different from what nature any where unfolds to contemplation. The language in which he speaks, is particularly distinguished from that which occurs in reality; but setting aside the circumstances of its consisting of verse, and figurative expression, those marked peculiarities which characterize the diction of a poet, he may fashion his language with a frequent disregard to the minuter rules of grammar. Criticism alone assumes the right of restraining the licentiousness of poetical

ardour; but even from its dogmas, he possesses a right of appeal to the judgment, and the feelings.

Science, therefore, appears to constitute a standard, from which the poet may be generally said to depart, in taking those liberties which are justified by licence. But that science is exclusively the standard from which he deviates, will be more admissible, on demonstrating the improbability of there being any other.

On a casual view of the subject under consideration, art and nature may be thought to possess an equal or paramount claim to that of science, in forming the standard from which the poet possesses a liberty of deviating in his delineations. From the rank which his compositions hold, as the principal among the arts of taste, the intimacy which they possess to the former, may be thought even closer than that which they bear to the latter. Nay, in regarding poetry as strictly imitative, in which is inferred the notion of an original and a model, which the artist aims at copying or emulating, each may be regarded as forming a standard, to which he must in some respects conform, and from which he may in others occasionally depart.

With a reference to this distinction, the licences of poetry may be conceived to be determinable; either as deviations from that state of things, abstractedly considered, which obtains in nature, or from that mode of practice which is generally observed in art, not less by the poet himself, than by such artists as imitate the same objects with him.

A single observation, however, will be sufficient to shew, that neither of these principles can be taken as the foundation of a theory which will be adequate to define or illustrate that subtle quality of composition which I have undertaken to investigate. One example will, in fact, sufficiently evince that some licence may be used where there is no deviation from any such standard; beyond which circumstance, we need not seek any additional proof of the insufficiency of the principle under consideration.

For let us suppose, as a possible case, that the poet has occasion to represent some fact which history describes as improbably atrocious and unnatural, and that suitably to what was likely to occur, he describes it as merciful or upright; it is evident that in thus misrepresenting a known circumstance, he takes a liberty with truth, which is only jus

tifiable by licence; and yet in this process, so far may he be from deserting that standard which is assumed in the present hypothesis, that his conduct may be at once more conformable to nature, and more consistent with his general practice, as well as with that of every artist who may undertake to describe the same circumstance.

In a word, the apparent force of both hypotheses may be not merely explained away; but both may be reduced under the more comprehensive principle which was originally laid down, as being exposed to no similar objection. The fact is, that both art and nature may form constituent parts of science, the true standard by which every deviation is to be estimated which is admitted as a licence: for we have already seen, that what is generally prescribed in the former, gives rise to that system of rules, which constitute the laws of grammar and criticism; and that what generally obtains in the latter, furnishes history and physiology with their respective subjects. These sciences include no small portion of the materials of a poet; and it will probably be found, that it is only as each assumes a scientifick form, that it constitutes a standard, by which the liberties