What are the causes which render such a form desirable? In the first place, a circle, though in itself agreeable to the eye, is the most monotonous of all figures; there is no change in it-no commencement or termination-no point upon which the eye can rest with decision; the consequence of which is that an assemblage of circles is most fatiguing and wearisome to the eye; and has, in relation to groups of other figures, Very much the effect of a countenance utterly without character, and conversation altogether destitute of meaning, compared with marked features and vivid expression. Now, as it is generally very desirable to group pictures, the circle would, on this account, be a most disagreeable form: while the parallelogram admits of variety of form as well of size, according to the proportion of the sides, enters into simple and symmetrical groups harmonizes with the right lines of walls and roof, and saves a great dea of space. These, however, are only the upholsterer's reasons for preferring the parallelogram. The artist's are of far more weight. The first great inconvenience is that the line of sight, or horizon, must be the horizontal diameter, and this, as we shall presently see, would take away all power from the artist of indicating the elevation of the spectator; while perspective retiring lines would incline equally upwards and downwards, producing an artificial and disagreeable impression. And, in the second place, if, as is very often-we may say generally-the case, there be no positive, continuous, horizontal line in the picture, the eye, in the case of the circle, would have no criterion whatever whereby to judge of the rectitude of the verticals, it would be doubtful about its own position, and uncertain which lines it was to assume as horizontal. Nine times out of ten, therefore, the verticals would appear inclined, and the absence of the parallel terminating lines would thus be embarrassing to the artist, injurious to the drawing, and painful to the spectator. And, lastly, the laws of composition, as far as relating to shade and colour, are very much facilitated by a rectangular form; the portions of each can be much more accurately estimated and disposed than in the circle; and the scientific forms of grouping, pyramidal, cruciform, &c., become much more evident, and, therefore, much more agreeable to the spectator. Hence it appears that the circle is practically offensive, though scientifically true; and, therefore, that if we can, by any modification of design, turn it into the parallelogram, without infringing any law of vision, it will be a most important and valuable alteration.