with peeping rocks, large mossy stones, and all their soft and brilliant reflections, that the eye lingers upon them: the two banks seem as it were to protract their meeting, and to form their junction insensibly, they so blend and unite with each other. In Mr. Brown's naked canals, no¬ thing detains the eye a moment; and the two bare sharp extremities appear to cut into each other. If in such productions. a near approach to mathematical exactness were a merit instead of a defect, the sweeps of Mr. Brown's water would be admirable; for many of them seem not to have been formed by degrees with the spade, but scooped out at once by an immense iron crescent, which after cutting out the indented part on one side, was applied to the opposite side, and then reversed to make the sweeps; so that in each sweep the indented and the projecting parts, if they could be shoved together, would fit like the pieces of a dissected map*.

* When I speak of Mr. Brown's artificial water, I in-* clude without much scruple, the greater part of what has

Where these serpentine canals are made, if there happen to be any sudden breaks or inequalities in the ground; any thickets or bushes; any thing, in short, that might cover the rawness and formality of new work-instead of taking advantage of such accidents, all must be made level and bare; and, by a strange perversion of terms, stripping nature stark-naked, is called dressing her.

A piece of stagnant water, with that thin, uniform, grassy edge which always remains after the operation of levelling, is much more like a temporary overflowing in a meadow or pasture, than what it professes to imitate-a lake or a river: for the principal distinction between the outline of such an

been made since his time: I consider him as the Hercules to whom the labours of the lesser heroes are to be attributed, and they have had no difficulty in copying his model exactly. Natural rivers, indeed, can only be imitated by the eye. either in painting or reality; but his may be surveyed, and an exact plan taken of them by admeasurement; and though such a representation would not accord with a Claude or a Gaspar, it might with great propriety be hung: up with a map of the demesne.

overflowing, and that of a permanent piece of water neither formed nor improved by art, is, that the flood-water is in general every where even with the grass, that there are no banks to it, nothing that appears firmly to contain it. In order, therefore, to impress on the whole of any artificial water a character of age, permanency, capacity, and above all, of naturalness as well as variety, some degree of height and of abruptness in the banks is required, and different degrees of both; some appearance of their having been in parts gradually worn and undermined by the successive action of rain and frost, and even by that of the water when put in motion by winds: for the banks of a mill-pond, which is proverbial for stillness, are generally undermined in parts by a succession of such accidental circumstances. All this diversity of rough broken ground, varying in height and form, and accompanied with projecting trees and bushes, will readily be acknow ledged to have more painter-like effects, than one bare, uniform slope of grass; that

acknowledgment is quite sufficient, and the objections, which are easily foreseen, are easily answered; for there are various ways in which rudeness may be corrected and disguised, as well as blended with what is smooth and polished, without destroying the marked character of nature on the one hand, or a dressed appearance on the other; of this I have already given some few instances*. But as artificial lakes and rivers are usually made, the water appears in every part so nearly on the same level with the land, and so totally without banks, that were it not for the regularity of the curves, a stranger might often suppose that when dry weather came the flood would go off, and the meadow be restored to its natural state. Sometimes, however, it happens, that the bottoms of meadows and pas tures subject to floods, are in parts bounded by natural banks against which the water lies, where it takes a very natural and varied form, and might easily from many points, and those not distant, be

• Vide my Letter to Mr. Repton, page 142.

mistaken for part of a river: to such overflowings I of course do not mean to allude, the comparison would do a great deal too much honour to those pieces of water, the banks of which had been formed by Mr. Brown; for it is impossible to see any part of them without knowing them to be artificial.

Among the various ways in which the present style of artificial water has been defended, certain passages from the poets have been quoted*, to shew that it is a great beauty in a river to have the water close to the edge of the grass:

May thy brimmed waves for this
Their full tribute never miss.

Vivo de pumice fontes

Roscida mobilibus lambebant gramina rivis*.

To which might be added the well known passage:

Without o'erflowing full.

* Essay on Design in Gardening, page 208.

+ Claudian de raptu Proserpinæ.