Can we Trust Palladio? Antoine Desgodetz Details Palladio’s Inaccuracies
Classical Comments by Calder Loth
Senior Architectural Historian for the Virginia Department of Historic Resources. Member of the Institute of Classical Classical Architecture & Art‘s Advisory Council.
Andrea Palladio’s Book Four of I Quattro Libri dell’Architettura (1570) is devoted to his documentation and reconstruction drawings of numerous Roman temples based on his extensive study of the ruins. He stated his intent for this endeavor in the introduction to Book Four as follows:
I intend therefore to illustrate in this book the form and ornaments of many ancient temples of which one can still see the ruins and which I have recorded in drawings, so that anyone can understand the form and ornaments. . . . and although one can see only portions of some of them standing above ground, I have nonetheless proceeded to deduce from them what they must have been like when they were complete. . .
This was an extraordinary effort. It marked the first concentrated recording project involving above-ground archaeological remains. It is mainly through these illustrations that we have any idea of the probable appearance of much of ancient Roman architecture. Moreover, the images became an important resource for classical design henceforth. Yet a close look at some of the details reveals that Palladio was not always accurate in his depictions of many features. We can speculate that Palladio relied on assistants to do some of the actual measuring while he wrote down the notes and made sketches based on the information being conveyed to him.
These discrepancies became clearly evident a century later when a young French architect, Antoine Babuty Desgodetz (1653-1728), was commissioned by Jean-Baptist Colbert, the King’s minister of finance, to travel to Rome to study the ruins in order to produce an architectural source book for French practitioners. Desgodetz had noted that Serlio, Palladio, and other Renaissance treatise writers were not necessarily reliable in their details. His determination to achieve a more precise documentation of the architecture of antiquity is noted in his own words.
For I found the means during sixteen months I was in Rome, to draw with my own hand those ancient structures, of which I have given the plans, elevations and profiles, with all the measures, which I have exactly taken. . . I have verified the whole over and over, in order to obtain a certainty for which I could answer. . . 
Desgodetz’s extensive fieldwork culminated in the 1682 publication of Les edifices antiques de Rome, dessinés et mesurés très exactement, a monumental and beautifully illustrated work on Roman ruins that has remained reliable for the precision of its documentation to the present. In his narrative descriptions of the ruins, Desgodetz bluntly described where Palladio missed the mark in his recordings.
The following is a sampling of instances where Desgodetz shows that Palladio frequently leaves much to be desired in the accuracy of his delineations and dimensions. The italicized quotes from Desgodetz are taken from a 1771 English translation of Desgodetz’s treatise by George Marshall, available in reprint by ECCO Print Editions.
Temple of Vesta, Rome (Fig. 1)
A storied landmark in the ancient Forum Boarium, near the banks of the Tiber, the so-called Temple of Vesta is one of the city of Rome’s few structures dating from the era of the Republic. In the form of a Greek tholos (circular temple) and constructed of Greek Pentelic marble, it likely was designed by a Greek architect and executed by Greek artisans. Contributing to this notion are the pointed tips of the abacuses, a treatment preferred by the Greeks for the Corinthian order, but almost never used by the Romans. Palladio missed this detail; his depiction of the capital is a standard Roman Corinthian version. Desgodetz lets us know of Palladio’s error, among others. (Figs. 2 & 3)
Palladio draws the capital quite otherwise than it is; he makes the bottom and top-leaves of one height, as they usually are, and puts five olive leaves to each division, he make the channels of the stalks twisted, ties the volutes together, and puts a small reversed leaf over them., he makes the volutes to ascend into the abacus, whereas they touch only the bottom of it, he puts a flower to bear the rose in the middle of the abacus, and makes not the rose as it is, he cuts off the angles of the abacus. . . .”
1. Temple of Vesta, Rome (Great Rift Valley Blogspot).
2. Palladio, Book 4, Chapter XIV, Temple of Vesta order [detail].
3. Desgodetz, Temple of Vesta order, Plate II [detail].
Temple of Fortuna Virilis (Fig. 4)
Next to the Temple of Vesta in the Forum Boarium is the diminutive Republic-era Temple of Portunus, the harbor god, which both Palladio and Desgodetz assumed was dedicated to Fortuna Virilis—manly virtue. Its richly decorated frieze was ornamented with swags alternating with candlesticks, bucrania (ox skulls), and putti. These frieze decorations are of stucco and have eroded over the centuries so that the only ornament remaining in situ is a single candlestick with short sections of garlands attached. (Fig. 5) Apparently enough of the frieze survived by the late 17th century for Desgodetz to record and make the following observation about Palladio’s illustration.
. . . he [Palladio] puts but one festoon between the children and the ox-heads, entirely omitting the candlesticks, and lays the festoons on the children’s shoulders whereas they bear them in their hands. . . He makes the festoons of the frieze with fruit, whereas they are of oak-leaves. (Figs. 6 & 7)
4. Temple of Fortuna Virilis, Rome (Loth).
5. Temple of Fortuna Virilis, frieze detail, (Loth).
6. Palladio, Book 4, Chapter XIII, Temple of Fortuna Virilis frieze [detail].
7. Desgodetz, Temple of Fortuna Virilis Plate IV: frieze [detail].
Temple of Antoninus and Faustina (Fig. 8)
Among the more conspicuous ruins in the Roman Forum is the Temple of Antoninus and Faustina. A defining feature of its entablature is the frieze decorated with carvings of griffins, Roman candlesticks, and acanthus foliage. (Fig. 9) These frieze decorations are found only on the side elevations. A dedicatory inscription occupies the frieze on the façade. (Fig. 10) As quoted below, Desgodetz was quick to note several discrepancies between Palladio’s plates and the temple as it actually exists.
Palladio puts on the front of the temple ornaments in the frieze, instead of the inscription. [In the frieze], Palladio has put no flowers, and turns up the griffins’ tails, which are trailing. He has placed directly over the middle of each intercolumnation the candlestick, on which the griffins have their feet supported, which is not the case, those candlesticks being set without order, and without any regard to the columns.(Figs. 11 & 12)
8. Temple of Antoninus and Faustina, Rome (Loth).
9. Temple of Antoninus and Faustina, frieze (Loth).
10. Desgodetz, Temple of Antoninus and Faustina, Plate II: elevation [detail].
11. Palladio, Book 4, Chapter IX, Temple of Antoninus and Faustina: capital and entablature.
12. Desgodetz, Temple of Antoninus and Faustina, Plate III: capital and frieze [detail].
Temple of Concord (Saturn) (Fig. 13)
The order employed in the columns of the Temple of Saturn,(which both Palladio and Desgodetz misidentified as the Temple of Concord, is one of ancient Rome’s rare examples of the Ionic order using angled volutes. (Fig. 14) A prominent landmark of the Roman Forum, the columns belong to a 283 A.D. restoration of a 42 B.C. temple damaged by fire. While Palladio’s illustration of the order captured its general character, Desgodetz’s more meticulous examination revealed many of Palladio’s small inaccuracies both in the capital and the entablature.
On the face of the capital he [Palladio] makes the volutes reenter the vase, he puts a quarter-round on the top of the abacus, where it is only a square list , he sets a flower in the middle of the abacus instead of a very peculiar ornament representing a chased  cup, he makes the volutes descend to the astragal of the top of the column, whereas they descend not so much as to the bottom of the astragal of the capital, he puts over each volute a leaf turned up, instead of a little scroll turned down . . . He puts imperfect darts between the eggs, instead of the double flowers that are there. . . . The form of the modillions is too small, he makes them answer directly to the middle of the columns, whereas a rose answers to it. (Figs. 15 & 16)
13. Temple of Concord (Saturn) portico, Rome (Loth).
14. Temple of Concord (Saturn) capitals (Loth).
15. Palladio, Book 4, Chapter XXX, Temple of Concord (Saturn): portico order.
16. Desgodetz, Temple of Concord (Saturn) Plate II: portico order.
Temple of Bacchus (Fig. 17)
The circular structure that both Palladio and Desgodetz identified as the Temple of Bacchus we now know to be the mausoleum of Constantina, elder daughter of the Emperor Constantine, built for herself around A.D. 351. It remains a well-preserved example of Late Roman architecture. A dominant feature of the interior is a colonnade of twelve pairs of Composite order columns supporting entablature blocks. (Fig. 18) As we note below, Desgodetz details the many inconsistencies in Palladio’s image of the order.
Palladio puts olive-leaves on the capital, instead of the leaves of acanthus, formed like parsley-leaves that are there, he does not make the flower in the middle of the abacus where it is, only a little foliage adorning the inside of the band which forms the revolutions of the volute. He omits the little hollow which is at the top of the abacus. On the architrave he makes all the bands plumb, whereas they project more below than above. He puts a hollow at the bottom of the cornice, where is an ogee, and a drip under the chin of the corona which is not there. . . . The whole capital is too low by twelve parts, the abacus alone by two parts, the architrave and frieze are both too low, the former by a part and a half, the latter by one part, the cornice is too high by a part and a half. (Figs. 19 & 20)
17. Temple of Bacchus (Mausoleum of Constantina), Rome (Italian-architecture).
18. Temple of Bacchus (Mausoleum of Constantina), interior Composite order (pinterest.com).
19. Palladio, Book 4, Chapter XXI, Temple of Bacchus, interior Composite order [detail].
20. Desgodetz, Temple of Bacchus, Plate III: interior Composite order.
Pantheon: (Fig. 21)
Circling the main body of the Pantheon are identical cornices composed of block modillions—short projections undercut with a cyma recta curve. These cornices have served as the model for numerous simplified modillion cornices, some of which are found in American colonial buildings. Being so visually accessible, we would think that Palladio would have no trouble accurately recording these cornices. Nevertheless, the indefatigable Desgodetz shows us that Palladio didn’t get them just right.
Palladio, in the profile he gives the second cornice, which goes round the outside, puts not the cymatium that is over the modillions, and places mouldings below, which are not there. He makes the height of the cornice too great by three parts. (Figs. 22 & 23)
Temple of Vesta at Tivoli (Fig. 24)
The romantically situated circular temple at Tivoli, long identified as the Temple of Vesta, but more likely dedicated to Hercules, is noted for its distinctive version of the Corinthian order. Its capitals have two compressed rows of acanthus leaves and an oversized fleuron resembling the hibiscus flower. The entablature frieze is decorated with fruit swags and bucrania. The latter are intact heads with flesh and eyes. (Fig. 25) Surprisingly, Palladio’s image of the order shows a generic Corinthian capital, and bucrania skulls, inconsistencies with what are actually there. Desgodetz describes these along with other discrepancies.
. . . he [Palladio] terminates semicircularly the flutings, which are all straight, and has expressed none of extraordinary particulars of the capital. He makes the second band of the architrave higher than that first, and joins the frieze by a curve to the list of the top of the architrave. He draws the ox-heads in the frieze without skin, and so give only the bones. He puts not a fillet that is at the top of the frieze, nor makes the drip low enough at the bottom of the corona of the cornice. . . . He makes the architrave too high by nine parts and a half, the frieze too low by one part, and cornice too high by fourteen parts and a half. (Figs. 26 & 27)
24. Temple of Vesta, Tivoli (LPLT/Wikimedia Commons).
25. Temple of Vesta, Tivoli: entablature detail (modeknit.com).
26. Palladio, Book 4, Chapter XXIII Temple of Vesta, Tivoli: Corinthian order and frieze decorations [detail].
27. Desgodetz, Temple of Vesta, Tivoli, Plate II: Corinthian order and frieze decorations.
The above notes and quotes are a random sampling of the many instances where Desgodetz takes exception to Palladio’s plates in Book 4. This essay is not intended to disparage Palladio and his invaluable pioneering work in documenting some of the great monuments of antiquity. His seductive reconstruction woodcuts of Roman temples have inspired architectural masterpieces from the Renaissance to the present. His illustrations capture the essence of the features if not the minute details. Desgodetz gave us precision, and helpfully contrasted his own findings with Palladio’s. Palladio’s Book Four can certainly continue to be a design resource for new classical works. However, if one wants exact reproductions of specific ancient details, Desgodetz is the source to use.
The details of the Palladian plates in this essay are reproduced from the Robert Tavenor and Richard Schofield 1997 translation of The Four Books on Architecture with the permission of the MIT Press.
Amanda Claridge, Rome, An Oxford Archaeological Guide (Oxford University Press, 1998).
Antoine Desgodetz, Les edifices Antiques de Rome dessinés et mesurés très exactment (Paris, 1682).
George Marshall, The Ancient Buildings of Rome by Antony Desgodetz Published in Two Volumes (London, 1761), ECCO Print Editions.
Andrea Palladio, The Four Books on Architecture, translated by Robert Tavenor and Richard Schofield (MIT Press, 1997).