fenfes language is the cause of the second, or any other fign that has the fame power with language; and a man's imagination is to himself the caufe of the third. It is fcarce neceffary to add, that an idea, originally of imagination, being convey'd to others by language or any other vehicle, becomes in the mind of thofe to whom it is convey'd an idea of the second kind; and again, that an idea of this kind, being afterward re called to the mind, becomes in that circum→ ftance an idea of memory.

20. Human nature is not fo constituted, as that its objects are perceived with indif ferency thefe, with very few exceptions, raife in us either pleasant or painful emotions. External objects, at the same time, appear in themselves agreeable or difagree

able; but with fome difference betwixt thofe which produce organic impreffions, and those which affect us from a diftance. When we touch a soft and smooth body, we have a pleasant feeling as at the place of contact; and this feeling we diftinguish not, at least not accurately, from the agreeablenefs of the body itself. The fame holds 3 C 2

1

in

in general with regard to all the organic impreffions. It is otherwife in hearing and feeing. A found is perceived as in itself agreeable; and, at the fame time, raifes in the hearer a pleasant emotion: an object of fight appears in itself agreeable; and, nat the fame time, raifes in the feer a pleafant emotion. Thefe are accurately diftinguishi ed. The pleasant emotion is felt as within the mind: the agreeableness of the object is placed upon the object, and is perceived as one of its qualities or properties. The a greeable appearance of an object of fight, is termed beauty; and the difagreeable appearance of fuch an object is termed uglinefs.

21. But though beauty and ugliness, in their proper and genuine fignification, are confined to objects of fight; yet in a more lax and figurative fignification, they are ap ply'd to objects of the other fenfes. They are fometimes apply'd even to abftract terms; for it is not unufual to say, a beautiful theorem, a beautiful conftitution of government. But I am inclined to think, that we are led to use such expreffion by conceiving

the

the thing as delineated upon paper, and as in fome fort an object of fight.

- 22. A line compofed by a precife rule, is perceived and faid to be regular. Aftraight line, a parabola, a hyperbola, the circumference of a circle, and of an ellipfe, are all of them regular lines. A figure compofed by a precife rule, is perceived and faid to be regular. Thus a circle, a fquare, a hexagon, an equilateral triangle, are regular fi gures, being compofed by a rule that determines the form of each. When the form of a line or of a figure is afcertained by a rule that leaves nothing arbitrary, the line and the figure are faid to be perfectly regular: this is the cafe of the figures now mentioned ; and it is the cafe of a straight line and of the circumference of a circle. A figure and a line are not perfectly regular where any part or circumftance is left arbitrary. A parallelogram and a rhomb are less regular than a fquare: the parallelogram is fubjected to no rule as to the length of fides, other than that the oppofite fides be equal: the rhomb is fubjected to no rule as to its angles, other than that the oppofite angles be equal.

equal. For the fame reason, the circumfe rence of an ellipfe, the form of which is fufceptible of much variety, is lefs regular than that of a circle.

23. Regularity, properly speaking, be longs, like beauty, to objects of fight like beauty, it is alfo apply'd figuratively to other objects. Thus we fay, a regular govern→ ment, a regular compofition of mufic, and, regular difcipline.

24. When two figures are composed of fimilar parts, they are faid to be uniform. Perfect uniformity is where the conftituent parts of two figures are precifely fimilar to each other. Thus two cubes of the fame dimensions are perfectly uniform in all their parts. An imperfect uniformity is, where the parts mutually correfpond, but without being precisely fimilar. The uniformity is imperfect betwixt two fquares or cubes of unequal dimenfions; and ftill more fo betwixt a fquare and a parallelogram.

25. Uniformity is alfo applicable to the constituent parts of the fame figure. The conftituent parts of a square are perfectly. uniform: its fides are equal and its angles

are

are equal. Wherein then differs regularity from uniformity? for a figure composed of fimilar or uniform parts muft undoubtedly be regular. Regularity is predicated of a figure confidered as a whole composed of refembling or uniform parts: uniformity again is predicated of these parts as related to each other by resemblance. We fay, a square is a regular, not an uniform figure: but with respect to the conftituent parts of a fquare, we fay not that they are regular, but that they are uniform.

1926. In things deftined for the fame use, as legs, arms, eyes, windows, fpoons, we expect uniformity. Proportion ought to govern parts intended for different uses. We require a certain proportion betwixt a leg and an arm; in the base, the shaft, the capital, of a pillar; and in the length, the breadth, the height, of a room. Some proportion is alfo required in different things intimately connected, as betwixt a dwelling-houfe, the garden, and the stables. But we require no proportion among things flightly connected, as betwixt the table a man writes on and the dog that follows him. Proportion

6