contributed in a great degree, and with great reason, to give a prejudice against the study of pictures as a preparation to, that of nature. In the same manner that many painters consider natural scenery merely with a reference to their own practice, many connoisseurs consider pictures merely with a reference to other pictures, as a school in which they may learn the routine of connoisseurship; that is, an acquaintance with the most prominent marks and peculiarities of differ ent masters: but they rarely look upon them in that point of view in which alone they can produce any real advantage,—as a school in which we may learn to enlarge, refine, and correct our ideas of nature, and in return, may qualify ourselves by this more liberal course of study, to be real judges of what is excellent in imitation This reflection may account for what otherwise seems quite unaccountable; namely, that many enthusiastic admirers and collectors of Claude, Poussin, &c. should have suffered professed improvers to

deprive the general and extended scenery of their places, of all that those painters would have most admired and copied.

The great object of our present inquiry seems to be, what is that mode of study which will best enable a man of a liberal and intelligent mind, to judge of the forms, colours, effects, and combinations of visible objects; to judge of them either as single compositions, which may be considered by themselves without reference to what surrounds them; or else as parts of scenery, the arrangement of which must be more or less regulated and restrained by what joins them, and the connection of which with the general scenery must be constantly attended to. Such knowledge and judgment comprehend the whole science of improvement with regard to its effect on the eye; and I believe can never be perfectly acquired, unless to the study of natural scenery, and of the various styles of gardening at different periods, the improver adds the theory at least of that art, the very essence of which is connec

tion: a principle of all others the most adapted to correct the chief defects of improvers. Connection is a principle always present to the painter's mind, if he deserve that name; and by the guidance of which he considers all sets of objects, whatever may be their character or boundaries, from the most extensive prospect to the most confined wood scene: neither referring every thing to the narrow limits of his canvas, nor despising what will not suit it, unless, indeed, the limits of his mind be equally narrow and contracted;. for when I speak of a painter, I mean an artist, not a mechanic.

Whatever minute and partial objections may be made to the study of pictures for the purpose of improvement, (many of which I have discussed in my letter to Mr. Repton,) yet certainly the great leading principles of the one art,—as general composition-grouping the separate parts-harmony of tints-unity of character, are equally applicable to the other: I may add also, what is so very essential

to the painter, though at first sight it seems hardly within the province of the improver-breadth and effect of light and

'shade.

These are called the principles of painting, because that art has pointed them out more clearly, by separating what was most striking and well combined, from the less interesting and scattered objects of geheral scenery: but they are in reality the general principles on which the effect of all visible objects must depend, and to which it must be referred.

Nothing can be more directly at war with all these principles, founded as they are in truth and in nature, than the present system of laying out grounds. A painter, or whoever views objects with a painter's eye, looks with indifference, if not with disgust, at the clumps, the belts, the made water, and the eternal smoothness and sameness of a finished place. An improver, on the other hand, considers these as the most perfect embellishments, as the last finishing touches that nature

can receive from art; and consequently must think the finest composition of Claude, whom I mention as the most ornamented of all the great masters, comparatively rude and imperfect; though he probably might allow, in Mr. Brown's phrase, that it had "capabilities."

No one, I believe, has yet been 'daring enough to improve a picture of Claude*,

so.

The account in Peregrine Pickle, of the gentleman who had improved Vandyke's portraits of his ancestors, used to strike me as rather outré; but I met with a similar instance some years ago, that makes it appear much less I was looking at a collection of pictures with Gainsborough; among the rest the housekeeper shewed us a portrait of her master, which she said was by Sir Joshua Reynolds: we both stared, for not only the touch and the colouring, but the whole style of the drapery and the general effect had no resemblance to his manner. Upon examining the housekeeper more particularly, we discovered that her master had had every thing but the facenot re-touched from the colours having faded-but totally changed, and newly composed as well as painted, by another, and, I need not add, an inferior hand.

Such a man would have felt as little scruple in making a Claude like his own place, as in making his own portrait like a scare-crow.