The Classical Orders of Architecture

Robert Chitham

• Publisher: Routledge
• Year: 2007
• ISBN:
  9781136358944
• Edition: 2
• Topic: Architektur
• Subtopic: Architektur Allgemein

This is the only publication that presents a modern interpretation of the Classical Orders. The new edition of this successful title now includes the proportions in both metric and imperial measurements to make the orders more accessible and to provide a valuable reference for designers.The inclusion of both 100-part and 96-part systems of proportion is underpinned by an essay on James Gibbs - one of the 18th century authors of standardized proportioning systems - and his influence in America. Along with additional plates, this book gives a
clear introduction to those not familiar with the classical genre and is an easy to follow guide which assists architects, interior designers and conservators with the quality of their design.

Use of the Orders

Plate 61

The Column: Diminution and Fluting

It was early discovered that the shaft of a column appeared more graceful if it diminished in diameter from base to capital. Most Renaissance authorities agree that this diminution should amount to one sixth of the base diameter (though Perrault allows rather less). I have adopted as a near metric equivalent of this a diameter at the head of 0.85 times the base diameter. It is not always clear at precisely what point this uppermost diameter should be measured, but I have assumed throughout that it is most logical to imagine the column length ‘produced’ through the capital to its junction with the entablature. Hence the breadth of the entablature is directly determined at 0.85 times the base diameter, whereas if the upper diameter is measured below the capital the dimension of the entablature is left indeterminate.

Since the ratio of upper to lower diameter is constant, the degree of curvature will be more pronounced in the shorter orders than in the longer.

Plate 61(a) takes as an example the Ionic column, as it falls between the extremes, but the method of setting out the diminution is identical for all.

Whilst the Greeks applied a curved profile to the whole length of the column, in Renaissance examples it became customary for the bottom third of the column to remain undiminished, in other words a cylindrical section, and for the upper two thirds only to be convex in profile. The curvature is so slight and subtle that it does not seem to me necessary for this division into thirds to be absolutely precise. In my drawing I have included the ovolo and bead at the head of the column (not strictly components of the shaft) as a reminder that in the Ionic order shaft and capital somewhat confusingly overlap. A further curiosity which is worth noting is that whilst the bases of all the orders are 0.5 diameters in height, only in the Tuscan is the fillet terminating the foot of the shaft included in the base.

Part (a) therefore demonstrates how to set out the curve of the diminished part of the shaft. At a point one third above the foot a semi-circle is drawn. A perpendicular dropped from the head of the shaft meets the semi-circle at A. The height of the upper part of the shaft is divided into any convenient number of equal parts (in this case seven), and the segment AB is similarly divided into equal parts, from which further perpendiculars are drawn. Where they cut the horizontal divisions of the shaft points of curvature are established, the line connecting these points being the curved profile of the shaft.

Chambers describes an alternative method, which he ascribes to Vignola. This is shown in part (b). CD and AF are the base and head semi-diameters of the shaft. An arc of radius equal to CD is inscribed with centre A, intersecting the shaft centre-line at B and the line AB is produced to meet CD produced at E. Any convenient number of lines B1E, B2E, B3E etc. are drawn (not necessarily equally spaced) and lengths equal to CD measured along them to identify points A1, A2, A3 etc. These points lie on the diminishing profile of the shaft. The lines CD, AB are so short in relation to the column height that unless the scale of the drawing is very large it is quite difficult to establish point E with any accuracy by drawing. I calculate, however, that angle AEC, for a diminution of 0.85:1, is approximately 31°45’. Chambers illustrates a pleasing device (attributed to Blondel) for drawing a continuous curve of diminution without recourse to all this laborious construction. Unfortunately, Blondel’s device only works for a specific set of column dimensions, so one would need to construct a new instrument each time a fresh order was to be drawn.

What mathematical formula governs the curves produced by these methods I cannot say. Robertson²³ states that ‘the curve in Greek work is usually a continuous hyperbola, but the Romans used the parabola and other forms, and sometimes made two different curves meet’. Whilst the precise nature of the curve is clearly of consequence in setting out the shaft full size for the contractor, in preparing ordinary small scale design or record drawings I believe any smooth curve of appropriate radius is adequate.

Traditionally, draftsmen used ‘railway curves’ to draw arcs of great radius. The computer now performs the same function with ease.

I have deliberately used the term diminution rather than entasis to describe the curved profile of the column. It is not at all clear whether the two terms are exactly synonymous. The term entasis is derived from the Greek enteino, to stretch, and means a swelling or convexity. This conveys the impression that the greatest diameter of a column is at a height some way above its base. There is a puzzling claim, made by Cockerell in a letter to Smirke written in December 1814, to have discovered the entasis of Greek columns.²⁴ He cannot have been referring to simple diminution, since this was understood throughout the Renaissance. Although the measurements he describes seem somewhat ambiguous, he seems to be claiming to have found examples of columns in which the greatest diameter occurs above the base. (Robertson, on the other hand, states flatly that the profile ‘in which the diameter increases for about half the height, is unknown in Greek work, at least before Roman times’.²⁵)

Certainly, the practice of diminishing columns towards both head and (to a lesser extent) base, has persisted, and it was particularly prevalent in the nineteenth century. It may be that the term entasis should be applied only to this variation. The two diagrams in (c) illustrate these two variations of the process of diminution. If diminution towards the base is adopted, it should be very slight; if it is overdone the effect may be somewhat comical.

Part (e) indicates the setting out of fluting, again using the Ionic shaft as an example. The Greeks used elaborate compound curves for flutes, involving great problems in the accurate execution of the work. Renaissance flutes are generally approximately semi-circular in section, and leave a fillet between adjoining flutes in the proportion of 2:6 in width (part (f)). Ionic, Corinthian and Composite shafts generally have twenty-four flutes, though the Doric may have as few as twenty. Tuscan columns are not normally fluted. Fluting may be continuous throughout the height of the shaft, or it may be confined to the diminishing part. Alternatively, the lower part of the flute may be filled with a convex cabling, as shown in part (c) which may itself be plain or enriched in the form of a rope or ribbon.

61. The Column: Diminution And Fluting

23. Robertson, op. cit, p. 117.

24. David Watkin, The Life and Work of C. R. Cockerell, p. 17.

25. Robertson, op. cit., p. 116

Plate 62

Intercolumniation I: Tuscan and Ionic

The spacing of columns has a considerable effect on the overall appearance of the order. A celebrated example of the sombre, claustrophobic effect of very close column spacing is the mausoleum at Castle Howard, by Hawksmoor. At the other extreme, very widely spaced columns appear to defy the structural constraints of a trabeated system of construction.

Indeed, much of the convention concerning column spacing arises from the limitations of span of stone lintels. Vitruvius cites five different spacings to which he refers in terms of intercolumniation, or the space between shafts, rather than the distance between column centres. The two narrowest are pycnostyle (2½ diameters between centres) and systyle (3 diameters). These he criticises on the grounds that they cramp the entrances to the building, and throw the wall within into deep shadow due to their close spacing. Diastyle (4 diameters) implies a considerable lintel span, and araeostyle (5 diameters) demands the use of timber in the entablature, and produces a squat appearance. The spacing favoured by Vitruvius he terms eustyle. This has 3¼ spacing between centres generally, but at the central openings this is increased to 4 diameters.

In practice, if the dimensions of the orders I have described in detail are adopted, the column spacings are in some instances circumscribed by the geometry of the elements of the entablature. The Tuscan and Ionic orders suffer from no restraint of this kind; the Tuscan order has no repetitive ornament to dictate the spacing, and the only repetition in the Ionic is in the dentils, which are so small in scale that they can be adjusted as necessary to suit any intercolumniation.

In the case of the Tuscan, I have shown the effect of spacing columns at 3 diameter centres (systyle), 3¼ (eustyle), 4 (diastyle) and 5 (araeostyle). I do not think, however, that it is necessary to stick rigidly to these Vitruvian spacings, and I have shown Ionic columns spaced at 2½ (pycnostyle), 3, 3½ and 4½ diameters. Spacings much above 4½ diameters are not in my view appropriate to continuous colonnades. The wider spacings are suitable for porches and aedicular elements. Where an even wider opening needs to be spanned, single columns produce a flimsy effect and alternative forms such as coupled columns, piers or arches (q.v.) must be employed.

62. Intercolumniation I: Tuscan And Ionic

Plate 63

Intercolumniation II: Doric, Corinthian and Composite

The spacing of Doric columns is constrained by the proportions of the elements of the frieze. The height of the frieze is 0.75 diameters, and in it square metopes alternate with triglyphs 0.5 diameters in width. This pattern thus recurs at 1.25 centres. The triglyph provides no scope for flexibility in its width, and the appearance of the whole is seriously affected if the shape of the metopes is much distorted. Since each column should have a triglyph aligned centrally in the frieze above it, this restricts the column spacings to multiples of 1.25 diameters.

At the narrowest possible spacings of 1.25 diameters, some elements of the base and capital merge, and the columns are said to be coupled. At 2.5 diameters a suitable, though somewhat close, rhythm appropriate for a continuous colonnade is generated. 3.75 diameters is still suitable for continuous repetition, but 5 diameters (araeostyle) is again only really suitable for porches etc.

The case of the Corinthian and Composite orders is a little more complicated. In both these orders the cornice is set out so that intermediate consoles or modillions align with the column centres. The proportions of the cornice and the spacing of modillions are determined by the careful proportioning of all the components and in particular the disposition of the coffers in the soffit of the corona. Gibbs in his detail plates provides the best arrangement of modillions and coffers. Unfortunately the modillion spacing set by these detail drawings does not correspond with that shown in either his general plate of the orders, nor in his diagrams of intercolumniation. Chambers consistently sets the spacing of modillions in both o...