VITRUVIUS, BOOK IV, CHAPTER 6
On the doors of temples





Contents of this chapter

  1. Text and translation
  2. Comment
  3. Bibliography
  4. Illustration of the Cesariano edition
  5. Fragments of renaissance treatises about the Doric order
    1. Serlio
1. Ostiorum autem et eorum antepagmentorum in aedibus hae sunt rationes, uti primum constituantur quo genere sint futurae. Genera sunt enim thyromaton haec: doricum, ionicum, atticurges.
Horum symmetriae conspiciuntur his rationibus, uti corona summa, quae supra antepagmentum superius inponetur, aeque librata sit capitulis summis columnarum quae in pronao fuerint. Lumen autem hypaethri constituatur sic, uti quae altitudo aedis a pavimento ad lacunaria fuerit, dividatur in partes tres semis et ex eis duae partes (semis) lumini valvarum altitudine constituantur. Haec autem dividatur in partes XII et ex eis quinque et dimidia latitudo luminis fiat in imo. Et in summo contrahatur, si erit lumen ab imo ad sedecim pedis, antepagmenti III parte; XVI pedum ad XXV, superior luminis contrahatur antepagmenti parte IIII; si ab pedibus XXV ad XXX, summa pars contrahatur antepagmenti parte VIII. Reliqua, quo altiora erunt, ad perpendiculum videntur oportere conlocari.

Translation

1. The following are the rules for doorways to temples and their architraves. First we must determine of what style they are to be. For the styles of doorways are these: Doric, Ionic, Attic.
Of these (as concerns the doric) the proportions are found to be of the following character. The top of the cornice which is put above the upper architrave, is made level with the tops of the capitals of the columns which are in the pronaos. The opening of the doorway is to be so determined that the height of the temple from the paviment to the panels of the ceiling is to be divided into 3,5 parts, and of these 2(,5) in height are to be fixed for the opening of the folding doors. Let this in turn be divided into 12 parts and of these let 5,5 be the breadth of the opening at the bottom. Let it be diminished at the top a third of the width of the doorjamb, if the opening be not more than 16 feet high; if from 16 to 25 feet, let the upper part of the opening be contracted ¼ of the doorjamb; from 25 to 30 feet, 1/8 of the doorjambs. Higher openings should have perpendicular sides.

2. Ipsa autem antepagmenta contrahuntur in summo suae crassitudinis XIIII parte. Supercilii crassitudo, quanta antepagmentorum in summa parte erit crassitudo. Cymatium faciundum est antepagmenti parte sexta; proiectura autem, quanta est eius crassitudo. Sculpendum est cymatium lesbium cum astragalo. Supra cymatium quod erit in supercilio, conlocandum est hyperthyrum crassitudine supercilii, et in eo scalpendum est cymatium doricum, astragalum lesbium sima scalptura. Corona plana cum cymatio; proiectura autem eius erit quanta altitudo. Supercilii, quod supra antepagmenta inponitur, dextra atque sinistra proiecturae sic sunt faciundae, uti crepidines excurrant et in ungue ipso cymatio coniungantur.

Translation

2. The doorjambs themselves are to be contracted 1/14 part of their width at the top. The height of the lintel is to be the same as that of the doorjamb at the top. The cymatium should be made one sixth of the doorjamb, projecting the amount of its thickness. It is to be carved in the Lesbian form with an astragal. Above the cymatium of the lintel, the frieze is to be placed as deep as the lintel; and on it is to be carved a Doric cymatium and a Lesbian astragal in low relief. Over this the cornice is to be carved without ornament and with a cymatium; its projection is to be as much as its height. To the right and left of the lintel, which is placed above the jambs, projections are to be made so that the bases run out, and are exactly joined to the cymatium with a mitre.

3. Sin autem ionico genere futura erunt, lumen altum ad eundem modum quemadmodum in doricis fieri videtur. Latitudo constituatur, ut altitudo dividatur in partes duas et dimidiam, eiusque partis unius ima luminis fiat latitudo. Contracturae ita uti in doricis. Crassitudo antepagmentorum (ex) altitudine luminis in fronte XIIII parte, cymatium huius crassitudinis sexta. Reliqua pars praeter cymatium dividitur in partes XII. Harum trium pria corsa fiat cum astragalo, secunda quattuor, tertia quinque et eae aeque corsae cum astragalis circumcurrant.

Translation

3. If the doorways are to be of the Ionic style, the opening must be of a height determined as in the Doric style. Let the breadth be determined so that the height is divided into 2,5 parts, and let the breadth of the opening at the bottom be one part. The contractions are to be as in the Doric. The width of the doorjamb is to be 1/14 of the height of the opening in front; the cymatium is to be 1/6 of the width of the doorjamb. The remainder, excluding the cymatium, is divided into 12 parts. The first fascia with the astragal is to be three parts of these; the second, of four parts; the third, of five parts. The fasciae with the astragal are to run evenly round the doorjambs.

4. Hyperthyra autem ad eundem modum componantur quemadmodum in doricis pro ratis pedibus. Ancones, sive parotides vocantur, excalpta dextra ac sinistra praependeant ad imi supercilii libramentum, praeter folium. Eae habeant in fronte crassitudinem ex antepagmenti tribus partibus, in imo quarta parta graciliore quam superiora.
Fores ita compingantur, uti scapi cardinales sint ex latitudine luminis totius XII parte. Inter duos scapos tympana ex XII partibus habeant ternas partes.

Translation

4. the tops of the doorways are as in the Doric style, with the proportionate dimensions. Brackets (or, as they are called, parotides) are to be carved right and left and to hang over to the level of the bottom of the lintel, with the leaf below that level. Their width on the face is two thirds of the doorjamb, being one fourth more slender at the bottom than the upper parts.
the doors are to be so put together that the hinge stiles are 1/12 of the breadth of the whole opening. The panels between the two stiles are to have 3 parts out of 12 in width.

5. Inpagibus distributiones ita fient, uti divisis altitudinibus in partes V duae superiori, tres inferiori designentur. Super medium medii inpages conlocentur, ex reliquis alii in summo, alii in imo compingantur. Altitudo inpagis fiat tympani tertia parte, cymatium sexta parte inpagis. Scaporum latitudines inpagis dimidia parte. Scapi, qui sunt secundum antepagmentum, dimidium inpagis constituantur. Sin autem valvatae erunt, altitudines ita menabunt, in latitudinem adiciatur amplius foris latitudo. Si quadriforis futura est, altitudo adiciatur.

Translation

5. For the rails, the distribution shall be such that, taking the height to be of 5 parts, two are designed to the upper portion, 3 to the lower portion. Let the middle rails be placed above the centre; of the others one set are at the top of the doors, the others at the bottom. The width of the rail is to be one third of the panel; the cymatium 1/6 of the rail. The breadth of the inner stiles is to be half the rail, and the covermoulding 2/3 of the rail. If the doors are folding, the heights remain the same; but let the breadth of the opening be increased. If the doors are fourfold, let the height be increased.

6. Atticurge autem isdem rationibus perficiuntur, quibus dorica. Praeterea corsae sub cymatiis in antepagmentis circumdatur, quae ita distribui debent, uti antepagmenti praeter cymatium ex partibus VII habeant duas partes. Ipsaqua non fiunt clathrata neque bifora sed valvata, et aperturas habent in exteriores partes.
Quas rationes aedium sacrarum in formationibus oporteat fieri (doricis), ionicis corinthiisque operibus, quoad potui attingere, veluti legitimis moribus exposui. Nunc de tuscanicis dispositionibus, quemadmodum institui oporteat, dicam.

Translation

6. Attic doors are made of the same proportions as the Doric, the fasciae, however, are carried round under the cymatia in the doorjambs, and ought to be so arranged that of the doorjamb, exclusive of the cymatium, they have 2 parts out of 7. They are to be without latice-work and are not to have hinges folding inwards, but to fold outwards on sockets.
As far as I could attain, I have set forth, as on approved lines, the methods which ought to be followed in planning temples of the Doric, Ionic and Corinthian orders. Now I will speak of the Tuscan style and the method to be employed therein.

COMMENT

The manuscripts of this text seem to be corrupt as far as it concerns the description of the Doric porch. Barbaro in his 1668 edition and Perrault in his 1684 translation (from which the above pictures are borrowed) give other proportions: the height from floor to ceiling is divided in 3,5 parts and of this 2, and not 2,5 as in our text, are given to the opening of the entrance. If we read this text carefully we come easily to the conclusion that 2 parts is to less because ’The top of the cornice which is put above the upper architrave, is made level with the tops of the capitals of the columns which are in the pronaos.’ This sentence led Barbaro and also Perrault in their reconstruction to put a very heavy and high corona above the frieze. But, this reconstruction is contradicted by other texts of Vitruvius where he uses the word corona to indicate a cornice. As illustration I can refer to book III, chapter V,13 where we find an enumeration of all the parts above the capital of the column, in the order they appear in the building: epistylia, zophora, coronae, tympana, fastigia, acroteria… architraves, friezes, cornices, tympana, pediments, acroteria…’. From this we can deduce that the corona is rather a small element and not the high rectangle which we find in the reconstructions of the 17th century. This rectangle is the logical consequence of the combination of a wrong measure (2 instead of 2,5) and the sentence that the top of the cornice is made level with the tops of the capitals of the columns.

If, on the other hand, we use the measure 2,5 everything comes to the right. This gives following reconstructions which I found in the Dutch translation  of 1999 (unfortunately, the tapering of the doors is omitted).

DORIC

IONIC

ATTIC

In the description of the contraction in the upper part of the door the basic proportion of the width of the antepagmentum (doorjamb) is missing. Vitruvius uses this measure to indicate the amount of contraction needed in accordance to the different heights of the doors. We find this measure only in sentence 3, when Vitruvius describes the Ionic door; the width of the jamb is here 1/14 of the height of the dooropening. Since the width of the cymatium is here the same as in the Doric porch (1/6 of the width of the doorjamb) we may assume that this proportion of 1/14 was also valid for the Doric door.

The construction of tapering doors was a common feature in antique architecture. We find it in Mycenean tombs, in Greek architecture (e.g. in the Siphnian treasury at Delphi or the Phileppeum at Olympia), in Etruscan graves, and, of course, in Roman architecture.

The best known example which is described by Alberti, Serlio, De l’Orme and Palladio is the so called temple of Vesta at Tivoli. Here we clearly can see the proportion of the dooropening as 2,5 portions of the 3,5 in which the height between floor and ceiling was divided. More, the top of the cornice is at the same level as the top of the capitals of the columns, just as Vitruvius prescribes. To illustrate this point I give the designs of Palladio from his book IV of his I quattro libri dell’architettura of the door and the windows of this temple and of the section of lintel, frieze and corona or cornice. This cornice is the upper protruding part of the construction. From this design it is obvious that the earlier reconstructions are absolutely impossible and that they are based on a false text.

Finally we must try to explain the reason for this tapering of doors and windows. Philibert De l’Orme in book VIII of his Architecture (1684) says in chapter V that the width of doors was lessened at the top because the lintels were made of one piece; so, the greater the span the greater the risk that these lintels could break under their own weight. To avoid this the jambs were sloping inward with the effect that the upper span of the door was reduced.
This explanation is maybe valid for the earliest examples but we can hardly imagine that it was effective in Vitruvius’ time. Indeed, when in sentence 1 he describes a door with a height of 16 feet (or 4,73 meter), the width at the bottom is 2,17 meter and the width of the doorjamb is 33,7 cm. If the contraction at the top is indeed only 1/3 of the width of the doorjamb, the span at the top will be still 2,06 meter. The effect of this contraction on the construction is negligible. Therefore I am inclined to think that this practice started indeed for constructional needs but that, after some centuries, it was purely seen as an aesthetic issue with an optical effect. To corroborate this assumption we must read the end of sentence 1. There we see that the higher the doors (and consequently the greater the width at the bottom) the lesser the upper contraction is made.
We can compare this practice to the practice of slightly sloping inward the outer columns of Greek peripteral temples. This was done to not give the impression that these columns were pushed outward. Despite the slight slope everybody who sees a Greek temple has the clear impression that the columns stay perpendicular to their stylobates.

When we try to place this text in Vitruvius' modular architectural system there are several difficulties. Maybe Vitruvius used a different source for this description of measuremants and proportions of doors. The only reference to the previously explained modular system of temples is the fact that the door opening stands in relation to the height of the temple from the floor tiles to the panels of the ceiling. This height shall be divided in 3,5 parts, 2,5 of which determine the height of the opening.
In doric temples the height from pavement to ceiling is 16,5 modules (1 module = width of a triglyph). It is impossible to divide this in 3,5 parts. The result (4,7142 modules) can't be converted into a practicable feet measure. The same goes for the Ionic temple where the height from pavement to ceiling equals 10 3/8 moduli (1 modulus = lower diameter of the column). It is also impossible to divide this in 3,5 parts because this gives an impossible measure (2,9642 modules) as well.

Bibliography

M. Vitruvii Pollionis de architectura libri decem cum commentariis Danielis Barbari, Venetiis, 1568.
Les dix livres d'architecture de Vitruve, Corrigés et traduits en 1684 par C. Perrault, Paris, 1684.
Vitruvius, De Architectura libri X, ed. F. Granger, London, 1962.
Ton Peters, Vitruvius, Handboek bouwkunde, Amsterdam, 1999.
Sebastiano Serlio, Tutte l’opere d’architettura et prospettiva, Venice, 1537-1554; translated in English by Vaughan Hart and Peter Hicks, London, 1996.
Philibert De l'Orme, Architecture, Rouen, 1648.
W.B. Dinsmoor, The architecture of ancient Greece, London, New York, 1927.
A. Boëthius – J.B. Ward-Perkins, Etruscan and Roman architecture, Harmondsworth, 1970.
H.Knell, Vitruvs Architekturtheorie, Darmstadt, 1985




Chapter 7
Back to contents Book IV
Home


Suggestions and remarks?
don't hesitate to send me a message