insist much upon the proportions of a column; and assign different proportions to the Doric, Ionic, and Corinthian. But no architect will maintain, that the most accurate proportions contribute more to use, than several that are less accurate and less agreeable. Neither will it be maintained, that the proportions assigned for the length breadth and height of rooms, tend to make them the more commodious. It appears then, so far as we can discover, that we have a taste for proportion independent altogether of utility. One thing indeed is certain, that any external object proportioned to our taste, is delightful. This furnishes a hint. May it not be thought a good final cause of proportion, that it contributes to our entertainment? The author of our nature has given many signal proofs, that this end is not below his care. And if so, why should we hesitate in assigning this as an additional final cause of regularity, and the other particulars above mentioned? We may be confirmed in this thought, by reflecting, that our taste, with respect to these, is not occasional or accidental, but uniform and universal, making an original branch of human nature. One might fill a volume with the effects that are produced by the endless combinations of the principles of beauty. I have room only for a slight specimen, confined to the simplest figures. A circle and a square are each of them perfectly regular, being equally confined to a precise form, and admitting not the slightest variation. A square however is less beautiful than a circle, because it is less simple. A circle has parts as well as a square; but its parts not being distinct like those of a square, it makes one entire impression; whereas the attention is divided among the sides and angles of a square. The effect of simplicity may be illustrated by another example. A square, though not more regular than a hexagon or octagon, is more beautiful than either; for what other reason, than that a square is more simple, and the attention less divided? This reasoning will appear still more solid when we consider any regular polygon of very many sides; for of such figure the mind can never have any distinct perception. Simplicity thus contributes to beauty. A square is more beautiful than a parallelogram. The former exceeds the latter in regularity and in uniformity of parts. But this holds with respect to intrinsic beauty only; for in many instances, utility comes in to cast the balance on the side of the parallelogram. This figure for the doors and windows of a dwelling-house, is preferred because of utility; and here we find the beauty of utility prevailing over that of regularity and uniformity.