good building, have failed miserably in the proportion of its parts; some of those we could name, were it not for the private injury of such remarks in the eyes of a discerning public. It is in this the ancient architects are discovered, by all that remains of their works, to have been most particularly excellent they formed at once an idea of the whole structure, both as to the internal parts as well as the external, and it is evident they throughout kept that general idea always in remembrance. It is thus we see such a perfect accordance in all their works; and from this, as we have shown in its proper place, is to be ascribed that harmony of parts; and these parts in the least of all their works are perfectly suited to one another. Now it is in this that the young architect, who would distinguish himself in his profession, should principally follow them in the disposition of a house. We err greatly, and the architect can scarcely set any modern model before him that is not defective; whereas, when he turns his eyes back to the ancients, there is not any one in which he will not find perfect truth and harmony prevail throughout.*

DISSERTATION XXV.

GENERAL REMARKS ON THE INTERNAL PROPORTIONS OF APARTMENTS.

"Proportion is the soul of beauty."-BURKE on the Sublime.

We have said, that a judicious distribution of the rooms, with easy access to each, is the essential required in a plan: we now come to the internal proportions of apartments: by this we mean that general order of the three dimensions which they possess; namely, length, breadth, and height, all which are essentially necessary to be observed to render rooms beautiful or pleasing in their form. Most people are able at first sight to say whether a room is well or ill-proportioned, although it is difficult to say what the principle is from which this propriety is determined. A room which has a low ceiling and is wide, is in general said to be heavy. A room, on the contrary, which is not too broad and has a high ceiling, is said to be light. The most common faults, accordingly, which observing people find with apartments, are either in their ceilings being too low, or the rooms too broad: the proportions of height and length they seldom attend to, if they are not greatly violated. The objection to broad ceilings is, that they are liable to sag in the middle, and therefore to crack the plastering; and where they are low, they will produce an oppression on the chest and a difficulty of breathing, which is not felt in high rooms.

To fix on any precise proportion, however, of length, height, and breadth for rooms, that would be a good general rule throughout a house with the greatest exactness, is impossible; for where floors are

* Andrew Borde, in his Dietarie of Helthe, has given us rules for planning and building a mansion-house according to the Tudor manner. He proposes a quadrangle, and directs the gate-house to be exactly opposite the porch of the hall; the privy-chamber to be annexed to the chamber of state, a parlour joining to the buttery or pantry at the lower end of the hall; the pastry-house and larder annexed to the kitchen. Many of the chambers to have a view into the chapel. In the outer quadrangle to be a stable, but only for horses of pleasure: the stable, dairy, and slaughter-house to be a quarter of a mile from the house. The moat to have a spring of water falling into it, and to be often scoured. An orchard of sundry fruits is convenient, but he rather recommends a garden filled with aromatic herbs. In the garden a pool or two for fish. A park filled with deer and rabbits: a dove-house also is a necessary thing about a 'mansyon-place; and among other thynges, a payre of buttes is a decent thynge about a mansyon. And otherwise for a great man necessary it is for to passe his tyme with bowles in an aley."

Too great a height in a room is not nearly so painful as too little height, and too great a length produces a trifling emotion of discontent, compared with that which we feel from too great a breadth. Whether a room is a few feet too high or too long few people observe; but every one notices a much less disproportion either in the diminution of its height or in the extent of its breadth.--(Alison on Taste.)

all of equal heights in rooms of different dimensions, how is this possible? Here arises a difficulty that will always occur, for if one room on a floor is made proportional, the others being of a less length and breadth must, of consequence, become disproportioned. Now there is but one method of remedy, and that, in many cases, is also liable to objection, I mean that of putting in a false ceiling,* which if made level would often bring the ceiling down below the heads of the windows: and, again, the same proportions which are beautiful in one apartment are not so in others. But there are limits to all proportions: to exemplify more clearly what we have said, we will take extremes; thus, for instance, a room twelve feet square may constitute a pleasing form, while a room five times as large, or sixty feet square, would be positively disagreeable. A room also twenty-four feet in length by eighteen in breadth is sufficiently pleasing, but seventy-two feet in length by fifty-four in breadth would constitute a very unpleasing form; and so on progressively in the same multiplying ratio. This must more or less be the case with all rooms on a floor, that are of the same height, and different sizes in length and breadth; here the small rooms are as high as the large ones; a disproportion in some of the rooms must therefore always prevail, unless, as we have said, false ceilings are put in to bring down the height in proportion to the width and length; a thing which is not usually done, and in most cases, as we have seen, impracticable, because the ceiling would then come down on the heads of the windows of the room, which windows are obliged to range with the front outside.

Another cause of this difference in the beauty and proportion of rooms seems to arise also from the character of the rooms themselves: every nice observer must here see that the several forms of rooms, their difference of magnitude, and various other causes, give them distinct characters, as to those of gaiety, simplicity, solemnity, grandeur, magnificence, &c. No room is ever beautiful which has not some such pleasing character; the terms by which we express this beauty are significant of the characters; and however regular the proportions of apartments may be, if they do not correspond to this general character we consider the form as defective or imperfect. Thus the same proportion of height which is beautiful in a room of gaiety or cheerfulness, would be felt as a defect in an apartment of which the character was severity or melancholy. The same proportion of length which is pleasing in an elegant or convenient room, would be a defect in an apartment of magnificence or splendour. The great proportion of breadth which suits a senate-house, and according with the severe and solemn character of the apartment itself, would be positively unpleasing in any room which was expressive of cheerfulness or lightness. In proportion also as apartments differ in size, different proportions become necessary in that respect, to accord with the character which the dif ference of magnitude produces. The same proportion of height which is pleasing in a cheerful room, would be too little for the baronial hall of a great castle, where vastness is necessary to agree with the sublimity of its character.+

Another cause of the difference of our opinion of the beauty of proportion, arises from the desti

* Where expense is not a hindrance, the height of the story may sometimes be suited to the principal rooms, and the middle-sized apartments be reduced by coving the ceilings with a flat in the middle, or by groins or domes, which will add to their beauty, independently of bettering their proportions. (J. S.)

The relation of breadth and height, which is so wonderfully effective in our Gothic cathedrals, although at variance with all the Greek and Roman classic rules of proportion, would be both absurd and painful in the forms of any common apartment. In general, I believe, it will be found that the great and positive beauty of apartments arises from their character; that when no character is discovered, the generality of people express little admiration even at most regular proportions; that every difference of character requires a corresponding difference in the composition of the dimensions, and that this demand is satisfied, or a beautiful form produced, only when the composition of the different proportions is such as to produce one pure and unmixed expression.—(Alison on Taste.)

nation of the apartment, and climate of the country; for instance, we should be better pleased with a large and lofty room in a warm country than in a cold one; and again, on the other hand, be better satisfied with a small room than a large one in a cold country. All apartments are intended for some particular use or purpose of human life; we demand, therefore, that the form of them should be according to those ends; and whenever the form is at variance with the end, however regular or generally beautiful the proportions may be, we are conscious of an emotion of dissatisfaction and discontent. The most obvious illustration of the dependence of the beauty of proportion on this species of utility, may be taken from the common system that natural taste has dictated, in the proportion of different apartments in great houses. The hall, the saloon, the diningroom, the drawing-room, the library, the chapel, the antechamber, the dressing-rooms, &c., have all different forms and different proportions. Change these proportions, give to the dining-room the proportions of the saloon, to the dressing-room those of the drawing-room, to the chapel the proportions of the antechamber, and every one will consider them as preposterous, unpleasing, and defective forms, because they are unfitted to the ends they are destined to serve. The most perfect beauty that the proportions of an apartment can exhibit, will be when every part is appropriate to its end, or when the same relation of dimensions which are productive of the expression of sufficiency, agree also in the preservation of character and in the indication of use.

DISSERTATION XXVI.

OF THE PROPORTIONS OF SEPARATE APARTMENTS.

"Symmetry is perfection itself."—VIOLA.

As proportion is one of the attributes of beauty, and the rules which guide it little understood, we shall proceed to develope it more at large. Now the proportions of apartments, as we have said, depend very much upon the use to which those apartments are devoted or intended to be applied. In general, however, it is to be observed, that the greater the capacities are, the more the length may exceed the breadth, and the smaller the house the more they may approach to the square. We have spoken of three sizes of rooms, the large, the middling, and the small, and of the necessity of each being formed on the plan in proportion to the whole edifice and to one another; but there yet remains a third kind of proportion to be considered, as well as that of the several divisions of the rooms, and here now comes in height. We have hitherto considered the room as a part of the house, and had every necessary regard to that as a whole; but we are now to suppose the architect to have drawn in the several apartments on his ground-plan, and is about to transfer them to the section and elevation, where the size or height is to be severally adjusted, and now remains no consideration but that of the respective measures for the purpose and the space each is to occupy. Now as these widths are not so certainly decided but that they may be a little altered if found necessary to the height, (one being made to give way to another,) the architect is to consider that he is not only to adjust the rooms to their lengths and breadths, but their height to both. Now this last is as nice a point for consideration as any in the whole compass of the science; nor is there any part of a building in which writers have established less certainty of rule. For the proportions of length

* Some persons are so fond of having a large drawing-room in their house, that all the other rooms must be made subservient. I remember to have seen a house of this description in Norfolk, and another in Kent, where a room in each was formed so large that all the others became comparatively closets.-(Author.)

and breadth on the plan are much easier to be ascertained than those of the height; for the former we have excellent rules in the works of the ancients, but in the latter we have none that has yet been established or laid down for absolute practice.

In the works of the ancients we find two general proportions of rooms, observed in respect of length and breadth; not that these were unexceptionable, for fancy sometimes exercised itself in this as in other instances, and none have indulged so much in it as those great architects. The two proportions, however, that were accounted most regular, and were most universal in their works, were, first, that one in which the length and breadth were made equal, forming a perfect square; secondly, that wherein the length was just twice the breadth, being one square in breadth and two in length. These were the standards of propriety in their square and oblong-square rooms; and we know by repeated experience that these are measures which suit very happily with one another. This we see, according to Vitruvius, was established into a kind of law.* What comes nearest to a certain proportion in their accounts with respect to height is, that in a room twice as long as wide the length and breadth should be summed up together, and the height then be set at half that measure ; thus a room of twenty feet long and ten wide was made fifteen feet high, which in this instance was, but for their large cornices and its being for a warm climate, much too high; however these seem to have been their most established proportions.

Since the Roman period the Italians have observed a variety of proportions, and many of them as happily. For dining-rooms some have divided the length into three parts, and given two to the width; that is, in the proportion of thirty by twenty, the width being divided into four parts; three of those parts were then given to the height, which example would be fifteen feet. For diningrooms some square the breadth, and then draw a diagonal line from corner to corner of the square, which they give as the best measure for the length and height; but this is excessive for height, though a good measure for the length of some rooms. For dining-rooms in a moderate-rate house one and a half the measure of the breadth is most proper for the length.

The manner of forming the ceiling naturally makes a variation in any given height of a room; we shall therefore recommend to the architect always to fix upon the form of this part before he calculates the height, and to vary that proportion accordingly. As flat ceilings are the most universal, we shall first notice them. Palladio, who is a good authority, lays down a rule for rooms upon the first-floor with flat ceilings, which is that their height be equal to the breadth from the floor to the joist. Upon the second story he varies the height by a considerable reduction; for he orders that it be made less than the breadth of the room by one-sixth, but when rooms on the first story have their ceilings arched, this author makes a difference, and with sufficient reason, for there he allows the height in a square room to be one-third more than the breadth. These proportions are all for large mansions and in warm climates.†

Now we see the ancients in square rooms allowed a greater height, and something is to be said in favour of each practice. These rooms were more magnificent and more light: we have instances of both proportions very exactly preserved in Italy, some in noble houses, built by Palladio, and others in such as are of an earlier date. It is not easy to determine which upon the whole is best, because there are evident advantages in each. The wisest method would be to endeavour at a proportion between the two, which should avoid the disadvantages of both. Such a middle proportion might

*Vitruvius, lib. vi. c. 5.

+ The breadth of the principal passages belonging to a moderate house may be one-fourth of the breadth of the principal rooms. The height of the passages should be the same as those of the rooms, but the lengths must be regulated by the building. To take off the heights in passages, they may be groined, or coved, or flat arched segmentally; and over the stairs the ceiling may be domed, having pendentives in the angles.—(Vignolo.)

have these, as well as addition of dignity, in imitation of the en suite rooms, and be light enough at the top, though not so light as those of Palladio's proportions. Whoever may set about to reform this eminent master, must remember that it is in England he builds, and not in Italy. The same space will not be so well lighted here by the same apertures as there, where the sky is clear and the sun brilliant; therefore, in establishing what we express by a proportion between the two, he is not to take the exact middle number, but rather something less than half the difference is to be added to Palladio's height. This is taking a mean between the two, though not in feet and inches, because that would be allowed in the difference of the Italian and English feet.

If this were all, the point were easily determined, but much more is necessary, between half and a third of the breadth added to the height: there may be innumerable proportions, the measures being minutely divided; of all these some one is indisputably better than another, but which that one is, has not yet been ascertained with any accuracy in those rooms with arched ceilings, whose length is greater than their breadth. Scammozzi, an Italian architect, directs a height to be formed, by adding the length and breadth, and dividing that measure in half. If a room of this proportion, for example, were twelve feet long and six broad, and had an arched ceiling, the height of it would be nine feet, that being the half of the joint measure of the length and breadth, which added together makes eighteen feet. This is a very plain and familiar direction, but it is further from exact truth and perfection than the other. This error is on the side of excess, not defect, and we advise the architect to endeavour to establish a proper height by repeating his variations, in making rooms of this proportion less and less by smaller degrees in height than according to the rules of Scammozzi.* The architect will now see that all he can learn of the three dimensions of length, breadth, and height from precept, particularly that of the latter, is but of little use; that a great deal is left to a good eye and a discretional judgment; and that there can be no certain law of height established which can universally and generally answer all the variety of dimensions of plan; it can only be acquired by practice and observation. He who has therefore seen and knows most, will be best enabled to decide and proceed with judgment. In rooms of the more usual dimensions, an architect will be at liberty to vary on every needful occasion, according to his fancy. This is very frequently necessary, but it is never done well unless by him who is master of all the settled proportions.

* As the height of rooms is of the greatest importance to health, as well as in an architectural point of view, this is necessary to be ascertained by some decisive rules. For an English climate, we may observe that the proportion of height is best to be taken from the width of the room alone, and not from the length; and that the principal room on the ground floor (which is to guide the subordinate ones) may be made in height equal to three-fourths of its width: supposing this room to be sixteen feet broad, the height would then be twelve feet. I have seen a room of the following dimensions, which struck me as well proportioned; it was sixteen feet broad, thirty-two feet long, and thirteen feet high. This gave to the height three-fourths and one-sixteenth part of its width.

As example is better than precept, we shall refer to a well-known noble-proportioned room in an ancient mansion, that of Hardwick Hall, in Derbyshire, belonging to the Duke of Devonshire. This once famous but now deserted house was commenced about the year 1592, and finished in 1597. It is in ruins, and save this one room, every wall is shattered and open to the winds of heaven. The building is gray with age, and almost overgrown with ivy, and threatening to tumble about the ears of the owls and bats, which are its sole inhabitants. Still here one majestic room remains entire, called the "Giant's Chamber," from two colossal figures in Roman armour which stand over the huge chimney-piece. This room has long been considered by architects as a perfect specimen of grand and beautiful proportion, and has been copied at Chatsworth and at Blenheim. The measures are as follows: length, fifty-five feet six inches; breadth, thirty feet six inches; height, twenty-four feet six inches; making the height about three-fourths and one-thirtieth part of the width.-(B.) In square rooms of the first-story, say some moderns, the height may be from four-fifths to five-sixths of the breadth of the side, and in oblong rooms the height may be equal to the width; this is too high, but an error in favour of height is preferable to making a room too low. The height of rooms in the second-story, it is further observed, may be one-twelfth part less than that of chambers below; and if there is a third-story, divide the height of the second into twelve equal parts, of which take nine for the height of those rooms.--(S.)