## The Five Orders

This plate represents my own distillation of the contents of the preceding plates comparing the orders by different authorities. First, an observation may be made about the manner in which it is set out. In preparing a drawing of this kind, two options are available: either the base diameter of each order can be kept constant, giving a stepped overall height, or the overall height of all the orders can be kept the same and a different module chosen for each. Renaissance writers are divided about this. Chambers is adamant in his opposition to making the module dimension constant: 'To render the comparison between the orders more easy, I have represented them all of the same height; by which means the gradual increase of delicacy and richness is easily perceivable . . . the proportions of the orders were by the ancients formed on those of the human body and consequently it could not be their intention to make a Corinthian column as thick and much taller than a Doric one.' He goes on to condemn a number of authorities, including Vignola, Scamozzi, Blondel and Perrault, for ascribing an ascending order of height to their orders.

Chambers' reasoning seems to be unacceptable; the comparison with the human body is a tenuous one, and we are seldom likely to use the orders to a human scale. I instinctively regard the orders as having a hierarchy of height, rather than of thickness. Besides, the advantage of setting them out to a common scale for the purposes of this plate needs no defence, as it is intended that the modular height of each will be grasped at a glance, by comparison with its fellows.

The actual proportions I determined upon as a result of comparing the ideal orders of the various Renaissance authorities stem from a number of factors. First, I accept the convention that the orders together form a kind of mathematical progression, from the Tuscan upwards, the column height increasing in steps of one diameter. I am not sufficiently bold to follow Perrault's logical conclusion of this progression and make the Composite taller than the Corinthian. In order to do this, one either has to make the Composite 11 diameters in height, producing an unacceptably attenuated appearance, cut the Tuscan back to Serlio's 6 diameters, or make the steps between successive orders less than i diameter as Perrault did. On the other hand, once the Composite is accepted as a variant of the Corinthian in terms of its proportions, the progression of column heights becomes very natural.

Second, I accept that the ratio of entablature to column height should be constant throughout the orders of my metric series. Not all authorities accept this. Gibbs, for example, adopts a ratio of 0.25:i for Tuscan and Doric and 0.2:1 for the remainder, in broad correspondence with Palladio who, however, has idiosyncratic ideas about the column heights. Chambers follows Vignola in adopting a common ratio of 0.25 :i, and it is this that I have followed on the grounds that it allows sufficient stature to the shaft without cramping the proportions of the mouldings of the entablature.

Having adopted a common ratio for the heights of the entablature and column, it seems logical to do the same for the pedestal which, perhaps because of its optional nature, has received much less attention in the past than the remainder of the vocabulary. Gibbs gives to the pedestal one fifth of the total height of the order, Vignola one third. Vignola's pedestals seem uncomfortably lofty; I have opted for a pedestal:column ratio of 0.3:1 throughout. I have been careful to apportion the pedestal height in a progressive way between its components of cornice, die and plinth so that as the die and plinth become more elongated with each successive order, at the same time the height given to cornice and plinth mouldings is increased to accommodate their greater complexity.

With the principal proportions now settled, the design of each order revolves about the selection of individual members either as a result of a general consensus between authorities, or in some cases merely through personal preference. Hence I have discarded devices which appear in only one set of orders, such as Serlio's gigantic bracketed Composite entablature, and similarly the modillion form of the Ionic published by Gibbs as an alternative form. I have followed Chambers' championship of the Corinthian as the culmination of the series, and afforded to it the most delicate and elaborate architrave, as well as reserving to it the curved form of the consoles supporting the corona, employing the much more robust square modillions in the cornice of the Composite. The Composite also is the only order to which I have allotted a pulvinated frieze, though a flat frieze would in fact serve here as a perfectly satisfactory alternative. I must express a preference, in the Ionic, for the symmetrical four-sided capital, which seems to me more shapely than its flat-sided counterpart. I have adhered throughout to the Attic base in preference to the Ionic, endorsing Chambers' contention that confusion is occasioned by joining several convex members together.

Beyond these general guidelines, the precise choice of mouldings and the sequence in which they are assembled must be left to the individual. It is the great strength of the orders that within a clear framework of leading proportions they offer scope for a wide variety of individual treatment.