Polzer AnatomyMasacciosHoly 1971

The Anatomy of Masaccio's Holy Trinity
Author(s): Joseph Polzer

Source: Jahrbuch der Berliner Museen , 1971, 13. Bd. (1971), pp. 18-59
Published by: Staatliche Museen zu Berlin -- Preußischer Kulturbesitz
Stable URL: https://www.jstor.org/stable/4125722
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
is collaborating with JSTOR to digitize, preserve and extend access to Jahrbuch der Berliner
Museen
This content downloaded from

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
All use subject to https://about.jstor.org/terms
THE ANATOMY
OF MASACCIO'S HOLY TRINITY
BY JOSEPH POLZER
Masaccio's Holy Trinity fresco in Santa Maria Novella has the double d
ing one of the most original and admired paintings of the early Renaiss
most travelled (fig. 1). Indeed, it is a miracle it has survived at all. Vas
then covered the fresco with his altar.' In the nineteenth century it
transferred to canvas and moved (in 1860) to the inside wall of the chu
Cavalcaselle, who had observed the fresco before and after this trans
praised it highly, commented on the damage which it sustained as a co
move.2 But the history of its travels does not end there. After the last war
found to be in such a perilous condition that it was thoroughly resto
Fortunately, recent advances in the techniques of restoration permitted
of its surface in that measure in which it can still be admired at this time. This arduous
restoration was thoroughly executed by Leonetto Tintori during the years 1950-54. He
also supervised the return voyage of the Trinity to its original location on the left aisle
wall in the third bay from the church entrance. Its exact location on this wall could be
fixed with considerable precision. The lower portion of the original fresco, depicting a
corpse lying on a sarcophagus, was found in situ just where it had been located by Antonio
Billi.3 The height of the fresco could be precisely fixed by a fragment of the original cor-
nice which was left behind on the wall at the time of the nineteenth century transfer. This
fragment included a plumb line made with a snapped rope which established the central
vertical axis. A lateral was then made with a snapped cord fixed at the level of the inter-
sections of two arcs drawn with a compass which was centered on the plumb line. This
lateral was independent of the level of the church floor. It extended along the lower termi-
nation of the egg frieze, where it can still be observed on the original fresco surface.
These features can be seen on a scale drawing after the fresco made by Leonetto Tintori
(see photo no. 68526 of the Soprintendenza alle Gallerie, Florence).
The experience of the Holy Trinity as it now appears is misleading since it gives the
false impression of completeness. During the nineteenth century transfer the missing
portions - and they include all the border areas - were painted in, and the recent restora-
tion respected this completion of the missing parts in the treatment of the area, largely
destroyed, below the Trinity panel. Closer inspection reveals many buckles and repairs,
surely scars of the many moves and injuries the fresco has suffered. It is indeed a marvel,
1 Vasari, Le Vite (ed. Milanesi, 1878), Vol. 2, p. 291.

2 Crowe and Cavalcaselle, Storia della pittura in Italia, Vol. II, 1885, p. 545.
3 I1 libro di Antonio Billi, ed. C. Frey, Berlin, 1892, p. 16.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
THE ANATOMY OF MASACCIO'S HOLY TRINITY 19
and a tribute to Tintori's excellence as a restorer, that the fresco h

power of attraction.
Before the recent cleaning in January of 1969 I was offered
examining the fresco at close proximity from a scaffold.4 Some obse
this examination I wish to share with the reader in the hope that
understanding of the fresco and its process of creation."
No preparatory studies have survived. To this writer it would seem
the ordered space, that the proportion and mass of architecture an
7Tinity should have been developed more or less directly on the wa
prior preparation. The very process of painting in buon fresco whic
ing from glornata to giornata, would exclude this procedure. Nor w
ever detailed, since it was covered by the plaster of the giornata to
ficed to guide the artist in sufficient detail. Masaccio must have develop
of the fresco slowly and meticulously. In the short time available
for the pigment to form a solid bond with the fresh plaster the artist
retical reflection. All the procedures of the giornata of painting h
ahead.
The Holy Trinity was painted in buon fresco. Twenty-four giornate can be distin-
guished from the top of the fresco to the bottom of the Trinity panel (fig. 5), around the
level of the step supporting the kneeling donors.6 In the normal sequence, Masaccio,
painted from top downward. As far as possible, his progress followed the architectural
setting around the figures. During the first eight giornate the monumental archway was
completed down to the level of the kneeling donors. The artist then painted most of the
barrel vault in the next two giornate, and only after this did he turn to the figures. The
reason for this procedure is simple enough. The symmetry and order of the composition
is defined by the archway in large measure. Once it is fixed the placement of the figures
poses no great problems.
Naturally, the artist must have begun by locating the dimensions of the fresco on the
wall. He then applied the plaster for the first giornata, which comprises the whole ent-
ablature and cornice extending from the pilaster capitals and the arch frame to the upper
limit of the fresco. The upper portion of the entablature is nearly wholly modern, with
4 I am deeply indebted to Leonetto Tintori for putting the scaffold at my disposal, and for his generous
advice. Eve Borsook, Howard Saalman, Roger Freemantle and Mark Zucker also lent their able assistance
in the taking of measurements. Harvard's Villa I Tatti offered the services of its able photographer, Luigi
Artini, who made many of the enlightening photographs in a raking light to bring out the surface
features.
5 The reader may also wish to consult Robert Oertel, "1Wandmalerei und Zeichnung in Italien", Mit-
teilungen des Kunsthistorischen Institutes in Florenz, Vol. V, 1940, pp. 507 ff.
6 The diagram of the giornate in figure 4 contains two alterations by the writer from the giornate
recorded on the diagram which was prepared by Leonetto Tintori in 1954 (photo: soprintendenza alle
gallerie Florence, no. 68526). The intrados of the arch frame is painted in a giornata after the latter, and
forms part, most probably (see ahead in text), of the giornata comprising the right ionic capital. In the
photo album he so kindly put at the disposal of this writer Tintori marked the separate giornate of the
face and the intrados of the arch frame on a photo of the Holy Trinity fresco before its restoration. In
addition, in the opinion of the writer the area below the right forearm of the Virgin comprising the open
sleeve belongs to giornata 17.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
20 JOSEPH POLZER
Fig. i. Masaccio, Holy Trinity fresc
the exception of an original segment

located at the center of the fresco, w
time of the last transfer. It included
ing arcs made with a compass resti
drawn through their intersections.
sions of the fresco had been fixed
giornata. The lateral rope line so
laterals above and below it. One of
establish the upper edge of the dentil

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
THE ANATOMY OF MASACCIO S HOLY TRINITY 21
Fig. 2. See fig. 1. Condition of fresco before restoration of 1950/54
In the first giornata Masaccio used a diversity of means to locate an

architectural detail. The dentil frieze was simplified so that the width
equal to that of the dentil. The vertical edges of the dentils were scr
indented in the plaster are clearly visible at the lower corners of the
faces. Toward the right side of the original fragment the scratched ve
diagonally downward toward center - ideally toward the vanishing poi
depicted the dentil in proper perspective recession. The indented point
the scratched verticals as they served to fix the sequence of the denti
lower limit.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
22 JOSEPH POLZER
The perspective treatment of the

The recessed central portion of the
cult to "read" as a coherent archit
with the assitance of a spolvero. I
vertical lines formed by fine points
a free brush work.7
During the second and third gior
same day - Masaccio painted the tw
destroyed, but the other is well pr
face missing at the lower right (fig.
It is interesting to note how Masa
a grid of rectangles by means of s
the plaster. The width of this grid i
below the capital. This grid must h
scratched roughly into position f
also scratched into the plaster, but
covered into three rows of superpo
they get wider in ascending order.
The upper portion of the capital
order of the grid. Masaccio shifted
ically to the right. This rightwar
capital from below and from the l
foreshortening, is compressed, and t
ing the center of the rosette is mov
The circle defining the contour of t
Masaccio's capital is very similar t
degli Innocenti (fig. 7).
The pattern of snapped rope lines
lower termination of the first giorn
of the entablature, is extended in f
during the third giornata. This ext
covering this capital. It follows that
antedates the giornate. The center
though it should have been situate
ward in corrected optical vision. M
formation of the capital he probab
to give the essential reference poi
the formation of the opposite capi
surely made free hand, since they ar
The fourth giornata comprises the
7 The use of spolvero for the formation

Orcagna (E. Borsook, The Mural Painters

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 3. See fig. 1. Diagram of fresco with the giornate indicated
by the ionic capitals, the intrados of the arch frame, and above by t
entablature (fig. 8). This entablature projects a short distance from
side of the entablature is shaded, and its width defined by a snapped
of the moulding of the arch frame are defined by a series of concentric
the surface. A central plumb line crosses the arch frame. Four inden
along this line, forming three equal intervals from the bottom, and
marking the juncture of the underside of the entablature with the
arcs are scratched through the three lower points they are unrelate
the arch frame which consists of three forward staggered strips top
this cyma extends a thin terminal strip which brings the arch frame
of the face of the entablature above, and below of the pilaster strip
ionic columns. All the edges of the mouldings are defined by arcs scratc
The two tondi set in the spandrils are of particular interest (figs.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
24 JOSEPH POLZER
Fig. 4. See fig. 1. Detail of first giornat
so situated that they touch on th

Their frames project to the face o
frame. Inside the tondi appears a sun
away from a central point at whic
cyma followed by a final projecting
The tondi are fascinating because
tion. It will be noted that the circles
concentric, but are made with a c
oriented toward the vanishing po
plaster and also the points marked o
This central point is level with the
point that the circle was made wh
the wall. Another point lies four-a
center for the outer circle of the
follow on the inside defining the ou
This point must have been disturb

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
the compass, and for this reason it was stuccoed over and retouched o
changes in color and texture of the pertinent areas with respect to the
clearly evident. Somewhat below this area is another point from wh
circle defining the bottom edge of the cyma. Beneath it, at a dis
three-quarter centimeters from the center of the rosette, another poin
lishing the circle which marks the termination of the fluting of the r
This termination, at which level the flutes are truncated along a pla
picture plane forming a sort of melon dome seen from the inside, is d
sequence of points marking the terminals of the ridges separating the
are determined by lines scratched free hand into the intonaco.
It can be observed that the same procedure was followed for both to
original surface allows confrontation. Their diameter is about 56 cen
Masaccio formed the grid for painting the right pilaster capital he alr
the location of the tondo. On the vertical defining the left side of th
indented point at 18 centimeters below the bottom edge of the entabl
radius of the tondo. The circumference of the tondo, if extended thro
volute of the overlapping portion of the pilaster capital, passes through th
During the fifth giornata Masaccio painted the ionic capital at the
Fig. 5. See fig. 1. Detail of first giornata: pseudo-meander with spolvero

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
26 JOSEPH POLZER
Fig. 6. See fig. 1. Detail of third gi

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
archway (fig. 11). Its scrutiny reveals indented points used for t
scheme of snapped rope lines. The basic element of this scheme
meters which is the width of the abacus and the height of the cap
shaft. At the left side this square conforms to the edge of the pil
giornata line) and at the top it is coextensive with the upper edge
right side and at the bottom the square crosses the surface of the cap
sharp angle from below. There is also a vertical rope line which sec
width of the arch frame at slightly more than
20 centimeters from the edge of the pilaster.
Indented points are evident at the termina-
tions of the rope lines. In addition, three in-
dented points fix the centers of the volutes and
of the knob situated midway between them.
The arcs of the volutes and the half circle of
the central knob were made with a compass

centered on these points. In contrast to these
precise lines one finds others roughly scratch-
ed into the plaster shaping the moulding of
the neck of the capital in its perspective re-
cession. The roughness of these lines shows
that they were made free hand.
An interesting feature on this giornata is
the relationship of the column shaft to the
capital (fig. 12). The column shaft was con-
ceived as about equal in width to the arch
frame at its juncture with the capital, about
Fig. 7. Brunelleschi,
20 centimeters. Masaccio had difficulty inOspedale degli Innocenti, pilaster capital,
shaping the elliptical optical path of the up- Florence
per edge of the column shaft. The segment of

the bottom side of the square covering the cap-
ital which extends from the side of the pilaster to the vertical connecting with the inside
of the arch frame is made into the upper side of a tilted square, oriented toward the van-
ishing point. This tilted square is abbreviated by the edge of the giornata but enough re-
mains to show that it followed the plane of the upper side of the column shaft and that
the elliptical path of the neck, its circular shape distorted by the optical angle, was in-
scribed in this square. A number of indented points and lines, ruled or drawn free hand,
are evident along the perimeter of the tilted square and of the column neck and in the
space between. Their quantity suggests that Masaccio was caught unprepared and search-
ed for a solution on the spot. The upper side of the tilted square was sectioned into six
equal parts, and so were probably the tilted sides. The center point of the upper side is
connected to the points marking the first interval on the tilted sides. The point marking
the first interval from the right onthe front side is connected to the point probably mark-
ing the center of the tilted side at the right, and a similar arrangement would correspond

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
28 JOSEPH POLZER
Fig. 8. See fig. 1. Central portion of a

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
on the opposite side of the tilted square. This system, extended to t

the tilted square, would form a tilted octagon inserted within thi
have served in turn for approximating the path of the ellipse of the
face (fig. 13).
The relationship of the pilaster edge to the column shaft is not q
stands free of this edge by 3 centimeters. It is 19 centimeters wide, bu
slightly more since it is recessed from the picture plane. Since one side
into which the column shaft is inserted is coextensive with the recedin
pier supporting the pilaster, the column would touch the wall. On the o
impression leaves a fringe of uncertainty whether the shaft is barely f
During the giornata which followed, the sixth, was painted the
right side of the archway (fig. 14), and also the intrados of the arc
servation of this capital reveals the same pattern of snapped rope lines
observed in the other: the square 31 centimeters long and 31.5 centi
tical descending at about 20.5 centimeters from the edge of the pila
in which the upper edge of the column shaft is inscribed is also present
tion visible within the area of the giornata (fig. 13). On the whole
ference points and rope lines evident. This capital was painted after
cio dismissed whatever marks he would have found superfluous. Th
the centers of the volute ends and of the central knob are presen
exception of the furthest, the compass was not used in forming the re
The study of the capitals yields within narrow limits their real di
of the column shafts below them. It can be observed that the egg friez
symmetrically centered inside of the vertical strip in the width of
slightly over 20 centimeters wide in the left capital and 20.5 in th
width of the capital at the level of the abacus would be the width o
the distance from the edge of this strip to the end of the abacus,
centimeters. The widths of the column shafts must be considered
actual readings amount to 19 centimeters for the left one, and 18
other. But it has been observed that they are partially recessed fro
Since they were inserted within a tilted square of sides equal to the
arch frame, their real width can be given accordingly as close to 20 cen
The intrados of the arch frame was painted during the same day
capital because the scratched arc defining its inside border extends into
of the capital. The seams of the giornata comprising the intrados ar
both long curving sides, and there is also to be observed a change in
tween its surface and that of the arch frame above it.
8 Masaccio reduced the number of scratched curves and applied them with greater care when he
articulated the moulding of the neck of the capital at the right side of the entrance arch. Also, judging
by the few points and lines Masaccio seemed more certain of how to establish the elliptical optical path
of the upper edge of the column shaft. A tilted square was drawn with its diagonals - visible within the
area of the giornata. Points are marked off on these diagonals so that the ellipse can be inscribed in such
a way that it touches on the midpoints of the sides of the tilted square and crosses the diagonals at the
above mentioned points. How he obtained their location is not clear.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
30 JOSEPH POLZER
Fig. 9. See fig. 1. Right tondo ov
The intrados was divided longitudi

(fig. 15). From one of these, situat
verted pyramids set on square bases
formed. Four arcs were made which l
lower sides along the front face of
visible to the beholder). On the line
their base, points were marked off
mids and of the spaces between them
the reversed tips of the pyramids, a

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 10. See fig. 1. Left tondo above arch
sides of the bases. With these marks the hanging pyramids could be eas
position.
During the two giornate which followed (the seventh and eighth) the
and pilasters were painted down to the kneeling donors. Accordingly, at
eighth giornata the vital figural portion yet to be painted and the most d
tive problem of the sharply receding vault seen di sotto in si' had been
archway. Masaccio then turned to the area within.
At this point perspective and its transfer from preparatory study to wall b
cial problem. How Masaccio went about this task can be observed partia

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
32 JOSEPH POLZER
Fig. 11. See fig. 1. Left ionic
giornata when he painted the le

giornata abounds with assisting
lines and indented points. During
the vault down to the cross arm h
and reduced his assisting scheme
In the ninth giornata the orthogo
sides and axis (fig. 16). Surely, the
ing point, but from points made

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 12. See fig. 1. Detail of juncture of left ionic capital with col
Fig. 13. Diagrams for Masaccio's schemes of approximating

the optical ellipse of the upper edge of the column in shafts
flanking the archway of his Trinity fresco

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
54 JOSEPH POLZER
Fig. 14. See fig. 1. Right ionic
this function are still visible, for in

gonal rib at the left of the center. T
points toward the opposite ends of
curving halfrib closing the vault in
also evident along the upper edge of
rib and at the right side of the secon
The curving transversal ribs were
and axis, and these were scratched
between the curving ribs are bise

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. f5. See fig. 1. Intrados of arch frame (det.)
scratched firmly into the plaster. The question of how Masaccio determined the center
of the arcs forming the curving ribs, essential to the perspective structure of the vaul
will be examined later. Whatever this procedure may have been, it was too elaborate
be solved during the very giornata of painting. Surely, the points marking these center
must have been fixed beforehand on a scale drawing and transferred thence onto the
wall. Since these points lay along the vertical axis in the area covered by the plaster o
later giornate no traces are evident.
There is one snapped rope line which does not relate to the curve of the vault directly
It extends from the upper edge of the abacus of the left ionic column laterally to the upper
edge of God the Father's halo (fig. 17). This line conforms to the radius of the outer cur
of the archway. But even so, it had no structural purpose for when it was made the arc
way had already been painted down to the inside edge of the intrados. It thus repeated
radius which was significant structurally only at an earlier moment in the painting of the
fresco. In all likelyhood this lateral formed part of a surface geometry (see ahead) whi
established the upper level of God the Father's location in the fresco.
There is also the problem of the indented points located in the fields of the recessed

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
36 JOSEPH POLZER
Fig. 16. See fig. 1. Detail of vault pai
coffers. Each recessed coffer face,

rosette barely visible at best, is joine
The perspective treatment of these cof
ficult and must have been worked ou
points served as guides for their op
situated immediately above God the Fa
at the lower left corner of the inner c
slightly below both lower corners of
edge of the left adjoining rib. A line

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 17. See fig. 1. Detail of vault painted during giornata 9
left side of the right adjoining rib. From the junctures of this line with t
diagonals were extended crossing the coffer so that they terminated at
the interior coffer face was crossed by the rib above it. Obviously, th
into place when the coffer frame was already established. The purpos
was to fix the center of the rosette: at their point of intersection. It is si
crossing diagonals appear only once in the entire vault. Most likely,
to locate the rosettes by hand after this experiment. In the two coffers w
the former Masaccio abbreviated his system. He marked on both coffer
the inner coffer face by a point located on the edge of the adjoining

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
58 JOSEPH POLZER
Fig. 18. See fig. 1. Upper right portio
marked the lower right corner of th

end of the inner coffer face along t
stroke descends from this point sla
follows. This overly impetuous str
coffer face.
A number of indented points lie alo

between the lowest ribs. This line
frames in their distorted optical
actly on the spots they would rela
implying that Masaccio corrected
the wall.
In the tenth giornata the quantity of observed assisting lines and points is reduced (fig.
18). Now the curving ribs are defined by just two lines marking their sides. Experimen-
tation with a string imitating a compass would show that the centers of the curving ribs
fall along a vertical shifted slightly to the right of the vertical axis used for drawing the
arcs of the curving ribs in the previous giornata. Like shifts in vertical axis can be ob-
served elsewhere in the fresco, for instance, as one moves from the archway (fourth gior-
nata) to the central rib of the vault (ninth giornata), or from it to the head of God the
Father (eleventh giornata). The plumb line which probably marked the vertical axis of

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
the fresco on the arriccio was only roughly followed on the successive
covered it.
Indented points of relevance are also to be observed in the tenth giornata. On the upper
edge of the curving rib connecting with the upper portion of God the Father's halo these
points locate the intersections with the orthogonal ribs. It seems curious why these should
have been fixed in the middle of the vault rather than at the extremities of the giornata.
Fig. 19. See fig. 1. Detail of vault painted during giornata 10 with rope impressions showing
Obviously, two points are necessary for snapping a rope line into place. Surely, the
vanishing point was not used as one of these points since its remoteness would have made
the procedure inconvenient. Besides, the scaffold on which the painter stood would have
been located between the vanishing point and the vault.
Interesting irregular rope marks are imprinted in relief on the intonaco of the vault
between God the Father's head and the ionic capital on the right side. It seems that the
artist's (or his assistant's) hand slipped when the lines for the sides of the third rib from
center were snapped into position. These lines were done over, one line to each side being
irregular, curving or tilting into the rib and then disappearing. Most probably the irre-
gular rope marks resulted from the whiplike action of a released taught line (fig. 19).
In this paper the scrutiny of the fresco is extended to those areas especially pertinent
for its perspective structure and the problem of the transfer from preparatory study to
wall. Presently, we move on to the left ionic capital situated at the rear of the vaulted

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
40 JOSEPH POLZER
Fig. 20. See fig. 1. Upper body of

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
space, painted during giornata 16 (fig. 20).9 It has been observed

from left to right and abbreviated the assisting lines and points w
matter identical with the exception of its reversal. For this reason
ing the ionic capital at the right side of the rear of the vaulted area.
Actually, the giornata comprising the left rear ionic capital includes
ital belonging to the side space compartment of the vaulted room whi
same pier. The giornata line begins some three centimeters above
level on down the surface of the giornata reveals a web of vertical and
The two verticals present are extensions of the sides of the arch fram
capital (identical to the respective verticals found on the ionic capi
of the vaulted room). The laterals define the width of the abacus; t
sideward to form the spiral of the volute; the width of the egg frieze
down to the column shaft. Two diagonal rope lines are also visible,
of the lower edge of the lowest orthogonal rib, the other crossing th
It will be observed that the ionic capital in the side space compartme
same laterals extending through the adjoining capital, with the ex
lowing the bottom edge of the abacus. However, to express the ca
ward projection its parts are raised so that its neck is level with the e
and its volute with the abacus of the other (it extends somewhat
this capital, and of the respective capital in the side space compart
space, is eliminated. The reason for the elimination can be discer
painted the vault down to these capitals in earlier giornate which i
these abaci he omitted them, not thinking sufficiently ahead, and he
to correct this error when this omission must have become eviden
tals at a later time. Probably, Masaccio paid little heed to this "c
view of their near hidden location.
We now turn to the Virgin Mary, painted during the seventeenth and eighteenth gior-
nate. The body of Christ Crucified had been completed when Masaccio turned to the
lower area of the archway comprising the Virgin and John the Evangelist. Each figure
was done in two giornate which were on the whole symmetrically distributed. During the
first giornata was painted the head and closer arm (fig. 21), and the remainder of her
body was added later. These giornate extended beyond the figure of Mary comprising the
space between the cross and the left side of the arch. A feature of capital importance is the
quadratura of snapped rope lines which covers the Virgin's body. To be precise there are
two quadrature of which the more detailed covers the portion of her body painted in the
first giornata, excluding just the lower section of the closer arm and its broad open sleeve.
The two quadrature are related in such a way that each square in the lower grid is section-
ed in the upper finer grid into sixteen square parts.
Only the Virgin possesses a grid in Masaccio's fresco and there can have been only one
reason for its execution: to assist the precise transfer of the figure from the preparatory
9 In figure 5 the ionic capitals at the rear of the vaulted space are assigned to the same giornata. Of
course, they might have been done as well on different days.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 21. See fig. 1. Virgin's head and arm

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
drawing to the wall. The evidence strongly supports this assertion. The
to the area around it - even to the bottom of the archway since the gri
and-a-half centimeters above the upper step. The finer grid was laid
conforming generally to her head. The terminal points used for mak
clearly visible along the two side verticals. Additional indented points at
within the grid locate key features: points along the silhouette of he
especially a point at the corner of her right eye. One vertical is omitted fr
vertical which would have crossed her nose and mouth. Masaccio surel
her face clear of an excessive assisting line.
The larger grid spread over her main body was extended from the
it in a later giornata - except for her closer arm where one witnesses th
in the first giornata. The laterals were snapped into place from points
cals of which one crosses the Virgin's advanced foot and the other her righ
For why only the Virgin demonstrates a grid in Masaccio's fresco o
his dramatic concept in the context of the early Renaissance. Within th
tion of late medieval crucifixions the Virgin's appearance is revolution
wilt in grief as does Giotto's Virgin Mary in the Padua Crucifixion. In
erect: severe, impassive, her temporal connection to the Crucifix unc
in the direction of the beholder is general and unfocused; her gesture
fied Son is made for the sake of the beholder's instruction. She is clearly r
natural symbolic role: the embodiment of Mater Ecclesia. Her normal
have been particularly striking for an audience accustomed to her int
concern toward her supplicants: witness, for example, her hand extend
Scrovegni who offers her his chapel in Giotto's Paduan Last Judgeme
Masaccio's Virgin's "reach" transcends individual relationship. It is un
beyond the donors kneeling below her, unaware of their presence, to en
present and future, who represents the Community of Christians. Th
Christian is presented to the Throne of Grace.'o
10 For the symbolism of Masaccio's Holy Trinity fresco and particularly of the T
recently, Otto von Simson's article: "Ueber die Bedeutung von Masaccio's Trinitaetsfr
Berliner Museen, 1966, pp. 119-159.
Masaccio's Virgin is closely related to the Virgin on a Crucifixion from S. Giusto
hibited in the recent fresco exhibition in Prato (Due secoli di pittura murale a Prato.
sinopie e graffiti dei secoli ziv e zv, catalogue by G. Marchini, Prato, September-N
fig. 8). The fresco belongs around 1400 and seems related to Agnolo Gaddi. Both th
toward the beholder and hand gesturing toward the Crucified Christ, and Saint Jo
Christ in sorrow with this hands clasped before his chest, recall the respective figur
Trinity. It is interesting that another fresco from Prato in the same exhibition, a Vi
Enthroned from the Tabernacolo del Ceppo, given to Niccol6 Gerini, depicts an abs
Child standing on his mother's lap and grasping her veil in a way strikingly similar
Child on his S. Giovenale triptych in the Uffizi (photo: Soprintendenza alle Gallerie,
Marchini dates the fresco around 1591 (op. cit., p. 52, fig. 4). Regardless of the vast d
the connection of these frescos to Masaccio would indicate that Masaccio borrowed
his close predecessors- be they minor artists active in the neighborhood of Florence,
of the later trecento whom these minor masters would have imitated.
For Masaccio's relationship to the trecento see also M. Boskovitz in Zeitschrift fuer Kunstgeschichte,
Vol. XXIX, 1966, pp. 51-66.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
44 JOSEPH POLZER
Obviously, Masaccio's fresco cann

merges the Holy Trinity with the
arymemorial.'l Its layered meanin
mass and weight of the figures; t
closure within Man's architecture
cipation", is made real in a differe
sotto in sz'. God the Father, Christ
Virgin's face is tilted in a calculate
- to realize the relationship of the
shoulders of Saint John also seem
filed face).
The grid used for her transfer from preparatory drawing to wall attests her special role
in Masaccio's fresco. He went out of his way to retain her precisely as planned.
From the observations thus far presented some conclusions can be reached. A compre-
hensive quadratura is not present. A spolvero is used only in one ornamental frieze in
the entablature. Architectural detail was located as needed by lines and points. The types
of assisting marks used changed from giornata to giornata. Straight lines were snapped
in with rope, circles and curves scratched in with the use of a compass, their locations
fixed with indented points made through scale drawings. Masaccio broke down the fresco
into its constituent giornate, and adapted his assisting marks to the requirements of each.
But Masaccio was aware of quadratura for purposes of transferring model to wall, as the
grid on the Virgin demonstrates. Thus far, no earlier grids are known in Italian paint-
ing. Two alternatives can be proposed for why Masaccio did not use quadratura for the
whole fresco: either this means of transfer was not yet known and practiced or he con-
sidered it unnecessary.
From a technical viewpoint Masaccio thus represents a transitional moment in the
history of fresco painting. Simplified conventions for the transfer or build-up of com-
positions on the wall which only approximated the shape and scale of component parts
did not meet the requirements of the Holy Trinity. The rigorous perspective of the archi-
tecture enveloping the composition imposed from the beginning fixed relationships of
shape and scale, and extended its order over the figures within. Masaccio's genial vision
of enclosing the Trinity within a real architecture conceived in a real space required new
modes of developing and transferring the composition to the wall so that these relation-
ships fixed by the perspective structure would be maintained without distortion. This ex-
cluded a primary study on the arriccio, hardly a suitable place for creating such a re-
11 The funerary character of the Holy Trinity is given by the connection of the donors with the
corpse of "Everyman" and the Throne of Grace. It remains uncertain, pending their satisfactory identi-
fication, whether they were acually alive or dead when the fresco was painted. John Pope-Hennessy has
observed that "initially the role of the Renaissance portrait was commemorative; it was consciously
directed to a future when the living would no longer be alive" (The Portrait in the Renaissance, New
York, 1966, p. 8). With the availability of death masks retrospective portraits of striking likeness would
have depended essentially on the artist' skill. Masaccio himself used the retrospective portrait when he
depicted Coluccio Salutati, who died in 1406, in the counselor seated below Theophilus in the Raising of
his Dead Son in the Brancacci Chapel (ibid., p. 7).

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
volutionary fresco. The Holy Trinity must have been bred and bo
"paper" in a situation of intense and sustained meditation, in a limited
lowed the artist to capture and project concepts in an initial stage of de
relevant that when Tintori stripped the lower portion of the fresco w
from the wall no traces of sinopie were uncovered.12
From the previous observations we must turn presently to the rationa
Masaccio's composition and space and relate our findings to those of o
these problems. These include primarily Kern's investigation of the pers
tion, written some fifty years ago,"3 and an interesting recent paper
which offers a scheme of proportions and measurements very tempting in
However, this logic is not quite as tight as Professor Janson's presentati
First we turn to the problem of dimensions. All dimensions here given
in situ more than once, unless otherwise commented.
A crucial fix on the fresco is the vanishing point. Kern located it a short
the step on which the donors kneel, which is situated about 180 centi
church floor.'" Considering the location to represent the normal eye lev
standing before the fresco, Janson places the vanishing point lower a
from the church floor. This level would put the vanishing point even with
"normal" quattrocento Florentine, smaller than his contemporary rela
the skeleton of "Everyman", 160 centimeters long, as the standard.16 H
servations bear out Kern's location.
Three participants and a rope were used to trace the paths of crucial orthogonals in
order to establish the area of their coming together. The orthogonals selected were the
fine axial rib lines snapped in with rope in the left half of the vault. The orthogonal ribs
of the vault were selected because they constitute the longest "depth" lines in the fresco.
Only the ribs in the left side of the vault were used because its construction preceded that
of the right side and seems more reliable, as has been stressed. Considering that these ortho-
gonals were probably transferred from a scale drawing the vanishing point - better: the
vanishing area - can be established with a surprising consistency in an area centering
close to five or six centimeters below the bottom step. Another vanishing point of un-
known manufacture can be observed on the modern wall (at this level the original sur-
face of the fresco is obliterated) at 8.5 centimeters below the bottom step, but it is too per-
fectly defined to sustain confidence. This point is still about 174 centimeters above the
church floor and considerably above Janson's. We conclude that "Everyman's" eye level
12 Verbal information offered by Leonetto Tintori. It leads to a reassessment of Oertel's insistence on

Masaccio's use of an elaborate sinopia (bibliography given in note 5).
13 G. F. Kern, "Das Dreifaltigkeitsfresko von Santa Maria Novella, eine perspektivisch-architektur-
geschichtliche Studie", Jahrbuch der koeniglich preussischen Kunstsammlungen, Vol. XXXIV, 1915, pp.
36-58.
14 H. W. Janson, "Ground Plan and Elevation in MIasaccio's Trinity Fresco", Essays in the History of
Art presented to Rudolf Wittkower, London, 1967, pp. 85-88.
Janson's article offers important corrections to Kern's Measurements (see note 15).
15 Kern, op. cit., pp. 59, 45.
16 Janson, op. cit., p. 85

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
46 JOSEPH POLZER
did not determine Masaccio's hori

or more.
We now turn to a second problem: whether Masaccio used stand

This has been proposed by Professor Janson who considers the palm
29.18 centimeters as the base measure."7 With some exceptions (see
ments taken in situ deviate substantially from Janson's half brac
cisive examples can be given. At the springing of the arch frame t
(equal to the distance between the pilasters) is not 204 centimeter
211 centimeters. Over this short span six or seven centimeters woul
able margin of error. Seven half braccie would come to 204.26 cent
can be confirmed by the vertical distance from God the Father's h
arch (inclusive of the arch frame). This distance measures 105.5 c
half that of the vault's width. This relationship is explained by the
is the radius and the latter the diameter of the same circle which
outer edge of the archway frame and defines the intersection of the v
ward, with the picture plane. Furthermore, the abaci of the ionic co
vault measure exactly 31 centimeters in length - nearly two who
than Janson's measurement of exactly one palmo. These examples
Nor do the schemes of assisting lines follow the palmo. The rope
capital to the right is somewhat more than 31 centimeters wide and 45
The tondi in the spandrils of the archway measure 35.5-36 centime
grid on the Virgin is composed of squares whose sides measure mos
meters (they vary from circa 12.9 to 13.2 centimeters). The front s
the vault - indeed, all sides if they be square, as seems probable t
measured indirectly. This can be done by extending the diverging orth
sides of the adjacent ribs (the moulded frames connecting the sunk
with the projecting faces of the ribs are considered part of the cof
intersection with the outer edge of the arch frame (which would co
section of the vault, if continued forward, with the picture plane).
tance between these points can be measured and would give the desi
can be adapted for obtaining the width of the ribs. The first area teste
row of coffers to the left of center. The lines extended forward were the axial rib lines
and the respective inner edges of the ribs touching on this coffer row. The results are inter-
esting. The coffer side comes to 28.5 centimeters, and the width of the rib to 11 centi-
meters. In the right half of the vault the lines extended involved the ribs bordering on the
second coffer row from the center. The lines chosen for extension were the inner edges of
the adjacent ribs and the furthest edge of the left rib. This time the coffer side came to
29 centimeters, and the width of the rib to 8.5 centimeters. The proximity of these coffer
sides to the half braccio is remarkable, and raises the possibility that Janson's theory
ought not to be completely discarded. Some irregularities in proportions might be ac-
counted for by lax transmissions of dimensions from cartoon to wall. But even an ex-
17 Ibid., pp. 84 ff.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
tremely permissive reading of our dimensions would not accomod

palmo scheme.
We now turn to another vital problem: the presence of a surfac
relationship to the perspective structure of the architecture. Professor
acutely that the receding vault, including the frame of the archway, c
in a square defined by the inner edges of the pilasters and extendin
lower edge of the entablature to the abacus of the ionic capitals s
vaulted space. Measurements bear out this observation, the height
207 and 206 centimeters at the left und right sides, respectively, a
centimeters at the level of God the Father's halo, and 211.8 centime
cross arm. The difference between height and width amounts to ci
margin would suggest doubt were it not that this surface geometr
Fig. 22 A. See fig. 1. Diagram of fresco with

surface geometry indicated

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
48 JOSEPH POLZER
second square would cover the low

centimeters, subtracting the height o
than 414 centimeters separating the
ture. The archway, including the he
angle composed of two superposed s
top step from the church floor (o
archway (211 centimeters at the lev
raised from the church floor. This
in Masaccio's surface arrangement o
The purpose of this surface geometry
method. Surely, it preceded the per
other than insertion within a squar
this geometry constituted a first d
of the architecture. This architect
imposing its realism and classicism
its gradual formation detail and pe
Relationships of surface geometry
the Father is defined at the top of th
upper surface square, and at the bo
in depth which curves to meet His
face relationships can be discerned,
and with it the arms and head of Chr
vertically from the very top down
served that the orthogonals extend
back and in front of the vaulted sp
the nails. These cupped fingers resp
tion. Furthermore, since the cross
tween the capitals (read from the r
might be suspended well inside the
pression is connected to the thesis
space which has found no few adv
phatically stressed.20 The Crucifix,
in closely connected planes. John c
supports God the Father, and the sh
greater height thus underscoring th
ship of depth and surface might h
it must have been retained as a sub
nal nature.
18 This measurement is approximate.

19 Kern, op. cit., p. 45; U. Schlegel, "Masaccio's Trinity Fresco in Santa Maria Novella", Art Bulletin,
March 1963, Vol. XLV, p. 25; R. Krautheimer and T. Krautheimer-Hess, Lorenzo Ghiberti, Princeton
Univ. Press, 1956, pp. 201, 245; P. Sanpaolesi, Brunelleschi, Milan, 1962, fig. opposite to p. 52.
20 Janson, op. cit., p. 88.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 23. Luca di Tommb,

Fig. 22. Lorenzo di Niccol6 Gerini,
Holy Trinity, panel painting,
Museo Nazionale di S. Martino, Pisa
Bob Jones University, Greenville
Of course, this rational nature remains the primary factor. It effec

without precedent. All figures are of the same family according to sca
God the Father being slightly taller (151 and 149 centimeters, respe
Virgin (13355 centimeters) and Saint John (129.5 centimeters). This
Masaccio's Trinity from those preceding his. Traditionally, God the
larger than his Crucified Son whose cross he touched or supported en
defined space. In a rare example of a standing (instead of seated) God t
di Niccol6 Gerini's 7rinity in Greenville, North Carolina (fig. 22), the
between Father and Son is especially marked. Only one rare and exce
Luca di Tomme's striking Holy Trinity in Pisa (fig. 23) - exhibits a c
which is as rigorously enforced as Masaccio's already in the sixties of t
the reduction of the sagging body of Christ to a lesser height than th
John's, as if the artist's measure had been the erect figure. God the F

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
50 JOSEPHI POLZER
scale appearing in form of a bust floa

Holy Trinity in Ss. Annunziata, wher
the Holy Spirit - are carried aloft by f
saccio saw and reflected on Luca's pain
Italian Renaissance painting preceding
However, scale is only one element in
or mortal, are enveloped in a measured
Father must stand immediately behind
above Christ, his feet are planted solid
nearly the whole width of the vaulted
to receive Him. This stringent subord
velous to behold just because He retain
ance, standing solidly planted on a gro
naissance artists returned God the Fat
Pesellino's in the National Gallery in L
bank,22 and in Raphael's Disput& whe
comprises God the Father who stands
waist by the arch which unifies the De
supplies a Trinity closest to Masaccio's
Fl6malle in Frankfort (fig. 25). There
standing on a base projecting from a
direction. But there the realism of the f
tures.
It stands to reason that Masaccio's determination to connect God the Father to the solid
ground determined in large measure the size and location of the box on whose projecting
lid He is standing. The level of God the Father's halo is determined, as has been seen, by
the diameter of the circle defining the intersection of the vault (if continued forward)
with the picture plane. His location and scale then determined in all likelyhood the height
of the box on which he stands. A discussion of the symbolism of this box, considered with
good reason as being the tomb of Christ,23 must take these factors affecting the shape of
the box into consideration. Whether this symbolism was intended from the start or
whether it evolved in the process of the formation of the composition will probably never
be known. In any case, it would support the essential funerary meaning of the fresco.
This fresco could well contain three tombs: below the Trinity that of "Everyman"; below
21 In a recent article Michael Mallory has stressed the close iconographic connection of Paolo di
Giovanni Fei's Holy Trinity - Crucifixion altarpiece in Naples Cathedral with Masaccio's fresco in S. Maria
Novella ("An early Quattrocento Trinity", Art Bulletin, Vol. XLVIII, 1966, pp. 88 ff.; see also M. Meiss,
Painting in Florence and Siena after the Black Death, Princeton Univ. Press, 1951, p. 55, n. 86; U. Schlegel,
in Art Bulletin, Vol. XLV, 1965, p. 25). I cannot accept Mallory's opinion in the light of the wholly
different formal and hieratic vocabulary of the Trinity by the Sienese painter which, different from Ma-
saccio's progressive realism, reverts to medieval hieratic and formal conventions.
22 See also Jacopo del Sellaio's Holy Trinity in a private collection in Munich (illustrated in Otto von
Simson, Ueber die Bedeutung von Masaccios Trinitaetsfresko", Jahrbuch der Berliner Museen, Vol. VIII,
1966, fig. 17 on p. 141).
23 See recently Janson, op. cit., p. 21.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
Fig. 24. Pesellino, Holy Trinity, panel painting, Fig. 25. Master
National of London
Gallery, Flemalle,
Staedelsches Kunstinstitut,
Frankfurt
God the Father that of Christ; in back of God the Father in the rear niche another - that
of the donors?24 This multiplication of sarcophagi holding diverse categories of deceased
would have its parallels, albeit in simplified form, in trecento funerary monuments. In
24 It can be observed on figure I that the rear wall of the chapel is not flat, but opens into a niche
whose juncture with the rear wall is clearly shown. The curving extension of this niche is depicted in
form of changing chiaroscuro along the curving surface of the wall. This niche is indicated in cross
section on Sanpaolesi's figure opposite page 52 of his book on Brunelleschi (bibl. in note 19). The entire
lower portion of this niche is closed off by an oblong stone slab, cut off partially by the lid on which
God the Father is standing. This stone slab is decorated along the edge by a thin molded rim, and its
central portion is hidden by the lower body of God the Father. The combination of niche with transversal
slab recalls the funerary niches so widely used in the trecento, for instance, along the very precinct wall of
S. Maria Novella. The molding of the rim of this slab recalls the moldings on the sides of the sarcophagi set
inside these niches. Often the sides of such trecento sarcophagi are composed of a series of three square
or rectangular fields, and it is probable that Masaccio imitated them. This sarcophagus was intended to
hold a corpse. If God the Father stands on the tomb of Christ, then it was not Christ's. It is possible then
that it was the burial place of the donors. The assumption would add a significant dimension to the
fresco's iconography. Below the crucifix lies "Adam-Everyman". His corpse would represent the beginning
of death's rule which extends from the fall of man to the end of the world. Of course, the Crucifix
symbolizes Christ's victory over death and the redemption of mankind. Could this redemption be depicted
extending to the donors by way of the superposition of the Holy Trinity: the Throne of Grace - over
their tomb?

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
52 JOSEPH POLZEl
Fig. 26. S. Croce, Florence, Bardi di Venio Chape

fresco of Entombment of Christ by Taddeo
the Bardi di Vernio chapel in S. Croce the wom

side of the northern wall reappears painted in sm
phagus of Christ in Taddeo Gaddi's fresco of t
ary niche (fig. 26). Her praying position reca
Trinity fresco.
Masaccio's reasoned nature does not exclude conservative or extraneous modes of cent-
ering attention on the Divine Personages. The same vault which recedes convincingly
into distance projects God the Father and Christ in the direction of the beholder. The
orthogonal ribs of the vault receding toward the vanishing point are like rays spreading
outward from the Divine, and the curving ribs echo in ever ascending arcs the scene with-

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
in of the Father supporting His Son. Physical movement inward and

radiating outward and toward the beholder are fused.25
Simultaneously, in contrast to the deep vaulted space the figures r
close to the picture plane, partially projecting forward. Witness, the
porting the skeleton of "Everyman", or the donors kneeling in fron
framing the Crucifixion-Trinity which is situated a short way within.
composition one notices how the figures, beginning with the skeleton,
as one's vision rises. The triangular formation of the figures assists th
ence to the foreground.
Analyzing Masaccio's integration of the figures within the architec
matter. However, since the receding vault is the most decisive spacial
fresco it deserved special attention.
We begin by questioning an assertion generally taken for granted: th
vaulted space is significant and that it can be measured.26 This has res
tionable assertions. Kern considered that the plan of the vaulted space
a consequence concluded that the coffers had to be rectangular.27 Di
wished to combine the square plan of the vaulted area with the square shap
and he believed this possible by extending the significant area to inc
space compartments which are only partially visible in the fresco.28
observed earlier how the ionic capitals flanking these side space com
abaci, and it seems unlikely that a significant space should have been
Indeed, the whole question of the significant, measurable plan of the
to be re-examined. Certain conclusions previously reached are import
shaped by insertion within a surface square. To all appearances, the nea
central area of the vault seem square, but the others are excessively
optical locations for drawing safe conclusions.
Additional information can be gleaned from further scrutiny of the
ing marks in giornata 9 where Masaccio considered the construction of
est detail have been described before. Some aspects of these assisting
reconsidered. In laying out the ribs two kinds of lines can be observ
scratched curves which barely dent the surface of the intonaco, and
which penetrate quite deeply. The rope lines defining the sides and c
25 Also Otto von Simson, bibliography in note 10, pp. 152.

26 See recently Janson, op. cit., p. 84. He refers to the importance of the square as
Renaissance humanistic speculation and religious architecture" following RP. VWit
Principles in the Age of Humanism, London, 1952, passim). The general validity of
mean that in Masaccio's case the notion of the square dimension of the vaulted area
ical scrutiny.
27 Kern, op. cit., p. 45.
28 Janson, op. cit., pp. 84 ff. Actually, Janson's calculation of the depth of the v
is not quite precise. Janson assumes that "the farthest band of coffers abuts direct
the 'rear entrance"'. But this is not correct. A rib half the normal width (one sixt
following Janson's calculation) joins the furthest coffer row to the arch frame of t
a similar half rib surely would correspond to it at the hidden front of the vault. It
rected depth of the vault, accepting Janson's thesis that each coffer row measure
intervening rib one third of a palmo, would amount to 91/3 palmi.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
54 JOSEPH POLZER
orthogonal ribs, and also the curves

versal ribs are finer than the more pe
and axis. This made them more distinct and contributed to the clear articulation of the
vault's recession. The presence in giornata 9 of all axial rib lines (we recall that they were
discarded in giornata 10) indicates that the coffering was defined originally by a plain
grid adapted to the curve of the vault which these lines would document. Accordingly, it
can be observed that the vault terminates at the back of the room with a rib half the
normal width - that is, at an axial rib line. The ribs and coffers would have been added as
a second step. Of course, this process of development would relate to the preparatory study
and not the fresco. Just why these axial lines were transferred to the fresco is not clear.
Possibly, Masaccio might have automatically transferred structural lines of significance
in a preparatory study but no longer necessary unless he considered that they assisted in
the development of the perspective in the giornata.
To all appearances, the vault comprises seven transversal rows of coffers, of which the
one closest to the beholder is hidden in back of the arch frame cutting off its vision. This
impression is generally acceptable, but it remains to be sustained by detailed observation.
Indeed, an interesting detail is relevant. Along the vertical axis the visible portion of the
vault comprises exactly six rows of coffers. Is this mere coincidence or part of Masaccio's
calculated design?
Several alternatives can be proposed. Masaccio might have formed the arch frame first,
and then accomodated a series of six transversal coffer rows within the visible vault; or
Masaccio might have deducted the optical width of the intrados of the arch frame down
from its apex along the vertical axis and deployed seven rows of coffers over the remain-
ing surface of the vault (partially hidden) - in which case the extension over the visible
portion of the vault of precisely six rows of coffers would seem a strange coincidence; or
again he might have formed the vault inserted within the surface square (as previously
noted) as composed of a grid of eight by eight coffer bands, then adapting the arch frame
to cover the two nearest ones.
The last alternative has the advantage of offering the simplest perspective construction
(fig. 27). The surface geometry of the fresco in its initial phase offered the square into
which the vault was to be inserted. The diameters of the hemicircles forming the front
and rear of the vault lay conveniently along the lateral bisecting the square and its bottom
side. The lengths of the radii were given by the lateral distance from the respective
centers lying on the vertical axis to the orthogonals defining the sides of the vault, set
along lines connecting the midpoints of the vertical sides of the square to the vanishing
point below. The outer hemicircle was divided into eight equal segments, and from the
seven points defining them lines were drawn to the vanishing point, these being the axial
lines of the orthogonal ribs. The spacing of the eight curving coffer rows was also a simple
matter. On the extended diameter of the hemicircle closing the vault in the foreground
at its intersection with the picture plane a sequence of eight equal segments of reason-
able length are marked off starting at center (experimentation shows that their lengths
are not fixed). From the end point of this sequence a line is drawn to the horizon so that
this line passes through the center of the hemicircle closing the vault in depth. Lines are

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
then extended from the intersection of this diagonal with the horizon to
points defining the segments of the above sequence. The intersections of
the vertical axis locate the centers of the hemicircles forming the axial li
seven transversal ribs, the radii being the lateral distances from these
orthogonals defining the sides of the vault.
Fig. 27. Diagram of Masaccio's initial perspective construction of the

coffered vault in the Trinity fresco, as proposed by the writer
For this perspective structure the measured depth of the vaulted space
necessary. Masaccio surely tested the result in that it corresponded empi
crete spacial image - and this impression has born the test from Vasari
present.
Let us observe how this alternative corresponds to the evidence. The pr
29 Vasari, Le Vite (ed. Milanesi, 1878), Vol. II, p. 291.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
56 JOSEPH POLZER
Fig. 28. Uccello, preparatory drawing f

Florence Cathedral, Uffizi, Flore
is whether the width of the arch fr

of the two nearest coffer bands whi
other which must be considered firs
band. This is no simple matter. Obse
exception of the nearest two visible
the rib. On the rib at the further side of the closest visible coffer row this axial line is
located well below center, and on the rib crossed by the intrados of the arch frame this

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
axial line is most probably the line located but half a centimeter ab
the rib (fig. 16).30 It can be observed that the surface of this rib ab
overlapping intrados, three centimeters wide, is clear of lines (the d
the edge of the giornata). Measurements bear out this assertion. Th
of the coffer rows from the rear of the vault forward is as follo
previous dimensions are approximate since along the central verti
both axial rib lines are covered by the head and halo of God the Fat
centimeters. Measured from the approximate centers of the respect
visible coffer band would measure about 19.5 centimeters - involvin
of nearly 5 centimeters over the width of the adjacent coffer row. It f
nearest ribs were raised slightly with respect to their axial lines w
drawn first. This correction would have connected the nearest coff
the edge of the intrados which lies at the line of the giornata. It is poss
transposition made this correction necessary.
We now turn to the next problem: whether the available space corresp
of the two replaced coffer bands. This space must be measured fr
axial rib line, situated, as has been seen, half a centimeter above t
rib. The pertinent distance extending from this line to the ape
amounts to about 45.7 centimeters. This distance would accomodat
such coffer bands - say of about 20.7 and 25 centimeters - to be joi
panding in the sequence of 11.8; 12.1; 14; 15.8; 16.6; 18.5 centimete
The solution here proposed for how Masaccio formed the perspect
the advantage of simplicity of design. The construction of the vault
logically from the surface square into which it was inserted. This w
in the two other alternate solutions previously proposed since the ce
hemicircle of the vault would have to be located empirically. After t
of the vault was fixed Masaccio then elaborated the archway in fu
frame and the columns which separate the vaulted space from that o
and the beholder. We have observed that the front face of the arch frame is of the same
width as the column shafts at its juncture with the ionic capitals. At the apex the front
face gains two centimeters in width. All the fluctuations in measurements cannot be ac-
counted for. Possibly, the increased width of the face assisted in extending the archway
and its intrados over the area of the two coffer bands which they replaced.
In surveying the previous discussion we can propose some general assertions. Masaccio
was obviously deeply bound to Brunelleschi in the shaping of his architecture and in his
concern and ability in perspective construction. But it would seem exaggerated to consider
the Trinity a thorough study in perspective on the same order of Brunelleschi's studies of
the Piazza della Signoria or the Piazza di S. Giovanni.31
The perspective of the vault is successful in the eye of the beholder, and it was formed
30 The measurements here given are taken from a tracing of the assisting lines and marks evident on
the vault.
31 Antonio di Tuccio Manetti, Filippo Brunellesco, ed. H. Holtzinger, Stuttgart, 1887, pp. 9-15.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
58 JOSEPH POLZER
by adapting a perspective easy to cons

of the fresco Masaccio emerges as a
system or detail. Some errors he left
the vision of the whole (we recall the
partments). The brushwork is broad
and there an inquisitive and restless
on the final intonaco which had no
scratched lines attempting to locate t
shafts at the sides of the arch.
In some measure it can be observed how Masaccio transferred the composition from
model to wall. A spolvero was used for just one architectural ornamental frieze of minor
importance. For the transfer of architectural shapes the artist made use of cartoons in the
size of the giornate. Through these cartoons essential points were indented on the into-
naco. These points were then connected to form schemes of lines sufficient for the detailed
reproduction of the objects to be painted. Occasionally, these schemes of lines were am-
plified by details scratched on the intonaco. A grid was used, but exceptionally for the
transfer of just one figure, albeit a vital one: the Virgin Mary.
In the development of fresco painting during the Renaissance Masaccio's Holy Trinity
plays a vital role. This fresco initiates a new phase of technical requirements. Before its
time a sinopia could accomodate the artist's control over the desired final statement. But
with the Trinity the traditional sinopia lost a good measure of its value. In its composition
all parts, figural and ambiental, were united in determined relationships of real appear-
ances and measurable proportions. These had to be transposed from the study to the wall
faithfully. This service the sinopia could not render and it is no wonder that no sinopia
has been found beneath Masaccio's fresco. On the other hand, the comprehensive grid is
yet missing or was consciously shelved. It will be recalled how insistently Alberti advocates
its use for the transfer of compositions,32 and how it appears on Uccello's drawing in the
Uffizi for his fresco of John Hawkwood in the Cathedral of Florence (fig. 28). In the
Presentation in the Temple by the Prato Master in the Cathedral of that city the mechan-
ical process of the transfer of a complex composition from preliminary study to the wall
by means of quadratura is clearly evident: there the grid appears as well on the grezzo as
on the intonaco levels.33
Masaccio's Holy Trinity has impressed its realism upon centuries of criticism. The
fresco testifies to Masacciols ingenuity in developing a vaulted space which is the shrine
of a humanised Divinity and an archway which is the threshold of Renaissance art.34
32 Leon Battista Alberti, On Painting, J. R. Spencer, translator and editor, Yale University Press,
1966, pp. 68 f.
33 Giuseppe Rosi, who restored the Presentation in the Temple from Prato Cathedral, was very gener-
ous with information. The appearance on the arriccio and intonaco levels of identical quadrature is
mentioned by Giuseppe Marchini in the recent catalogue of the exhibition of frescos in Prato (biblio-
graphy in note 10, p. 18, figures on pp. 101, 105).
34 The writer extends his gratitude to the Harvard Institute of Renaissance Studies at Villa I Tatti,
to its director Myron Gilmore, and to Millard Meiss, without whose interest and support this paper would
not have materialized.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00
PROVENANCE OF THE PHOTOGRAPHS
Soprintendenza alle Gallerie, Firenze: Fig. 1, 2, 5 (with some changes made by

28. - Luigi Artini, Villa I Tatti: Fig. 4-11, 14-18, 20, 21, 23. - The writer: Fi
versity, Greenville: Fig. 22. - National Gallery, London: Fig. 24. - Bildarchiv Mar
RESUMEE
Masaccios >>Trinitat<< in S. Maria Novella zu Florenz stellt nicht nur ein hervorrag
Strebens nach Naturwahrheit im friihen 15. Jahrhundert dar, sondern auch eine iibe
in der Entwicklung der Freskenmalerei. In diesem ersten Wandgemiilde der Renaissan
menschliche Figur und eine reiche, realistisch proportionierte wie auch tiefenriumlic
kiinstlerische Einheit bilden, hatte der Maler die entsprechenden Mittel erst zu erf
position von dem gewil3 kleinformatigen ersten Entwurf auf die Wand zu iibertrag
bestimmten Grd63enverhaltnisse und Details zu verandern.
Sorgfaltige Betrachtung der Oberflaiche des Freskos verrit, trotz vielfacher Besc
Kiinstler vorgegangen ist. Das Spolvero-System, d. h. die Verwendung des Pausverfah
Bildflaiche, so verbreitet es im spateren 15. Jahrhundert war, gibt es hier noch nich
nur in einem Detail an: bei der Gestaltung eines architektonischen Ornaments. Dass
Quadratura des ganzen Bildes, wie sie im spateren >>Tempelgang Mariae<< des Domes z
sie wird in der >Trinitat<< nur bei der Uibertragung der verkiirzten Muttergottes-Figu
Bei scharfer Seitenbeleuchtung erkennt man vielerlei Spuren an der Oberfliche, die
daBf Masaccio einen in entsprechende >>Giornate<< (Tagewerke) eingeteilten Karton
mui3ten wohl die Dimensionen und die vertikale Achse auf der Wand festgelegt we
wahrend der >Intonaco< (Feinputz) fuir die >Giornata<< noch feucht war, die Schliisse
indem man den Karton mit einem spitzen Werkzeug durchstach. Diese kleinen L6ch
wesentlichen Punkte und die Linien, die vorher durch Schniire angedeutet waren, f
auch die Mittelpunkte der Kreise zu finden, die fiir die Gestaltung der architekton
unerlfl3lich sind. Diese Kreise wurden dann mit einem Zirkel in den Feinputz einge
dienten als Grundlage fiir die architektonischen Formen wie Kapitelle, Friese, Siiulenge
Einzelheiten wurden hiufig freihindig in den Feinputz eingeritzt, so die gezackten Ri
bliitter an den Stiitzpfeilern, die Kapitelle oder die Kannelierung der Tondi. Masacci
wohl erfolgreich, erscheint deutlich selbst erfunden und beweist, daB die Technik der
von Kompositionen mit allen Details in der Florentiner Freskomalerei damals durch
iiblich war, obwohl man den >Spolvero<< (Pause) und die >Quadratura<< (Quadratnetz) k
Masaccios Tonnengew6lbe, dessen Perspektive di sotto in sit so iiberzeugend wirkt,
der linken Hiilfte, im Laufe von zwei >Giornate<< gemalt worden. Die erste >>Giorn
Linien der orthogonalen und der transversalen Rippen. Dies deutet darauf hin, daf
tierung auf der Grundlage eines Achsengitters entworfen wurde. Horst Janson hat
Wl61bung in ein Quadrat hineinpafBt, das die obere Halfte eines Hochrechtecks bildet
ebenfalls ein Quadrat ist. Der Rundbogen fiigt sich in seiner Breite in das obere Quad
punkt liegt in der Mitte der unteren Begrenzungslinie.
Masaccios perspektivischer Entwurf des Gew6lbes entstand innerhalb der Oberflii
perspektivisch verkiirzte Muster der Kassettierung konnte mittels einfacher Geometri
Masaccio benatigte weder einen TiefenmaBlstab fiir den gew6lbten Raum noch einen D
perspektivische Konstruktion war praktisch und wirkungsvoll. Die von Masaccio kon
fliiche besteht aus acht mal acht Kassettenreihen, von denen die zwei vordersten durc
mung und ihre Archivolte verdeckt werden.

91.242.155.167 on Thu, 30 May 2024 11:34:32 +00:00

Polzer AnatomyMasacciosHoly 1971

Uploaded by

Copyright:

Available Formats

Polzer AnatomyMasacciosHoly 1971

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Polzer AnatomyMasacciosHoly 1971

Uploaded by

Copyright:

Available Formats

The Anatomy of Masaccio's Holy Trinity

Author(s): Joseph Polzer

Stable URL: https://www.jstor.org/stable/4125722

This content downloaded from

OF MASACCIO'S HOLY TRINITY

1 Vasari, Le Vite (ed. Milanesi, 1878), Vol. 2, p. 291.

This content downloaded from

and a tribute to Tintori's excellence as a restorer, that the fresco h

This content downloaded from

Fig. i. Masaccio, Holy Trinity fresc

the exception of an original segment

This content downloaded from

Fig. 2. See fig. 1. Condition of fresco before restoration of 1950/54

In the first giornata Masaccio used a diversity of means to locate an

This content downloaded from

The perspective treatment of the

7 The use of spolvero for the formation

This content downloaded from

Fig. 3. See fig. 1. Diagram of fresco with the giornate indicated

This content downloaded from

Fig. 4. See fig. 1. Detail of first giornat

so situated that they touch on th

This content downloaded from

Fig. 5. See fig. 1. Detail of first giornata: pseudo-meander with spolvero

This content downloaded from

Fig. 6. See fig. 1. Detail of third gi

This content downloaded from

the central knob were made with a compass

per edge of the column shaft. The segment of

This content downloaded from

Fig. 8. See fig. 1. Central portion of a

This content downloaded from

on the opposite side of the tilted square. This system, extended to t

This content downloaded from

Fig. 9. See fig. 1. Right tondo ov

The intrados was divided longitudi

This content downloaded from

Fig. 10. See fig. 1. Left tondo above arch

This content downloaded from

Fig. 11. See fig. 1. Left ionic

giornata when he painted the le

This content downloaded from

Fig. 13. Diagrams for Masaccio's schemes of approximating

This content downloaded from

Fig. 14. See fig. 1. Right ionic

this function are still visible, for in

This content downloaded from

Fig. f5. See fig. 1. Intrados of arch frame (det.)

This content downloaded from

Fig. 16. See fig. 1. Detail of vault pai

coffers. Each recessed coffer face,

This content downloaded from

Fig. 17. See fig. 1. Detail of vault painted during giornata 9

This content downloaded from

Fig. 18. See fig. 1. Upper right portio

marked the lower right corner of th

A number of indented points lie alo

This content downloaded from

This content downloaded from