the rewards of service leonardo da vinci and the

!"#$!%&'()*+,-.*'/012('345(678.9()*+,-.*'"''"&:;:""''""9"&''5,<)'"#
the rewards
of service
leonardo
da vinci
and
the duke
of milan
LUKE SYSON
!"#$!%&'()*+,-.*'/012('345(678.9()*+,-.*'"''"&:;:""''""9"&''5,<)'"&
L
eonardo da vinci ’ s 18 years in milan
were the making of him.1 It was probably in 1482 that
he journeyed from the mercantile republic of Florence
to this, the wealthiest and most populous of Italy’s
dynastic city states, and he soon entered the orbit of its
magnificent ruler, Ludovico Maria Sforza (1452–1508;
nicknamed il Moro – ‘the Moor’ – probably because
of his swarthy features: see cat. 2). And when, in about
1489–90, Ludovico began paying Leonardo a salary, the
prince was granting the painter the time and space to
eΩect a quite extraordinary metamorphosis of the art of
painting. Leonardo’s three surviving Milanese portraits
(cats 5, 10, 17), not least his likeness of il Moro’s mistress,
Cecilia Gallerani, chronicle a stylistic journey that was
to revolutionise the genre.2 His two versions of the Virgin
of the Rocks (cats 31, 32) were both painted for an elite
Milanese confraternity, packed with Ludovico’s courtiers.
Superficially they look alike – their compositions are
more or less the same – but in their details and hence
their overall ambition they are revealed as profoundly
diΩerent from one another. These are disparities that
reflect Leonardo’s significant change of direction in the
years after 1490.
In 1550 Giorgio Vasari, the first great historian of art,
placed Leonardo in the vanguard of what he dubbed the
modern manner, notable for his ‘force and boldness of
design, the subtlest counterfeiting of all the minutiae of
Nature exactly as they are, with good rule, better order,
correct proportion, perfect design and divine grace’.3
Leonardo was being credited with the stylistic leap
that resulted in what many art historians call the High
Renaissance. This changed sense not just of what pictures
might look like, but of their whole purpose and scope, is
usually located in Florence during the years Leonardo
spent there immediately after 1500. But of all Leonardo’s
works Vasari devotes most attention to the Last Supper
(fig. 105), a work of almost uncanny perfection (despite
its rapid decay), executed in Milan in about 1492–7/8.
Quite properly, Vasari gives Ludovico Sforza a leading
role in the narrative of its execution. For it was actually
with this picture, and as Ludovico’s court painter, that
Leonardo had first attained that pioneering combination
of detailed naturalism, a feature already familiar from
the work of Netherlandish painters and their Italian
imitators, with something that is deemed new: the
‘divine grace’ that – with the artist seeking to surpass
the beauties of nature – could take painting into the
realm of the otherworldly.
In return for Ludovico’s protection, this marvellous,
modern painter would be celebrated as ‘his’, the human
emblem of the Sforza court. The rhetoric surrounding
his employment ensured that Leonardo’s highly visible
gifts were taken as the mirror of his patron’s more
abstract talents as a ruler. And, particularly in the 1490s,
Leonardo’s painting of a world made perfect by analysis,
discipline and imagination could be understood as corresponding to the much promoted notion of the prince
as the perfected ruler of an ideal state. Onlookers may
have been aware of the ways in which the life stories of
Ludovico and Leonardo chimed, making it clear that their
achievements were due to their outstanding talents, but
also the responsible ways in which they had honed these
gifts. Patron and painter were exactly of an age and their
roads to glory had been unconventional. Ludovico became
Duke of Milan only in 1494, but (with the title Duke of
Bari) had ruled the city as regent for his young nephew,
Gian Galeazzo, from 1481. However, as the fourth son
of Duke Francesco Sforza, he had been brought up
with no real expectations of power. Leonardo was born
with even fewer prospects, the illegitimate child of
a peasant girl and a middle-class notary, tucked away
in the Florentine countryside until his late teens, but
becoming a painter whose gifts were so manifest and
manifold as to guarantee his success. A publicly emblazoned
partnership between patron and painter could make their
contemporaries contemplate the question – of immense
rhetorical importance to both men – of where talent
comes to reside, of the diΩerence between a great man
and the rest.
The connections that can be traced between Leonardo’s
artistic trajectory and Ludovico’s rhetoric of rule should
emphatically not, however, be seen as matters of mere cause
and eΩect. Unlike many of his courtier contemporaries,
Leonardo was too creative and too independent to turn
himself into a servile panegyrist. And Ludovico was wise
enough not to attempt the complete annexation of his
painter’s immense creativity. Leonardo is often treated as
peerless, unconnected with the world around him, locked
away in the tower of his own genius. But his artistic
philosophy evolved against a background of collectively
13
!"#$!%%&'()*+,-)&./01'&234'567-8'()*+,-)&9&&9:;<;99&&9989"&&4+=(&>>
!"#$!%%&'()*+,-)&./01'&234'567-8'()*+,-)&9&&9:;<;99&&9989"&&4+=(&>%
f ig s 44, 45
Infrared reflectography details
of the London Virgin of the Rocks
(cat. 32)
f i g . 42
Diagram of hidden composition in
the London Virgin of the Rocks (cat. 32)
a lost original by Leonardo.44 Before the composition was
laid out on the panel Leonardo probably made similar
drawings for the heads of all the protagonists, as he did
for Saint John. Both provide early examples of Leonardo’s
lifelong habit of working out detailed elements within
larger compositions through the targeted use of highly
finished drawings, studies that may even have been
made after painting had begun, a practice which is most
famously evident in the next decade during the execution
of the Last Supper.45
The Louvre Virgin of the Rocks (cat. 31) is a significant
advance on Leonardo’s previous work (coming just after
the Adoration). For the first time we have a multi-figure
composition that is fully completed, set within a remarkably
complex landscape. Even in its compromised condition
the picture demonstrates a remarkable level of execution:
surviving details of hair, draperies and foliage display a
sustained degree of finish, equal to any of his portraits.
More important, each detail is carefully calibrated within
a larger scheme of relationships of colour, tone and illumination, which are coordinated across the whole of the
66
composition – an interest of increasing importance to
Leonardo, and one even more evident in the better-preserved
version of the painting now in the National Gallery.
The National Gallery Virgin of the Rocks (cat. 32) might
seem to be a simple reprise or variation of the Louvre
painting, but research undertaken in 2005 has made it
clear that its production was a more convoluted process
than its appearance suggests. Surprisingly, the panel was
begun with a wholly unrelated composition which appears
to be closely connected to compositional sketches for the
Adoration of the Christ Child at Windsor and in New York
(fig. 42, cats 30 and 40). The principal elements of the
kneeling Virgin of that initial composition are revealed
by infrared reflectography. Drawn with the brush in a
liquid medium, her head and left hand are based on some
sort of mechanical transfer from partial cartoons, while
the drawing of her drapery and right hand is much more
free and improvised, the latter in particular still sketchy
and unresolved (fig. 44).46
The relationship of this now hidden composition to
other works by Leonardo is of fundamental importance
for our understanding of the working practices of his
studio. The designs of the Virgin’s head and left hand
appear in other works, a clear indication of the existence
of partial cartoons, sometimes rescaled, for these features,
as well of as the reuse and recycling of such elements
f i g . 43
g i ova n n i a n ton i o b olt r a f f i o
and m a r c o d ’ o g g i on o
The Grifi Altarpiece (detail of fig. 98), about 1497
in pursuit of perfection
for a variety of purposes. The pose of the hand (fig. 44)
appears in the Last Supper (fig. 46), as well as in Portrait of
Cecilia Gallerani (cat. 10) and the Grifi Altarpiece (fig. 43) –
a documented work by Boltra≈o and Marco d’Oggiono
– both of which can be dated to the mid-1490s.47 The
underdrawn Virgin’s head (fig. 45) is extremely similar
to a reversed image of the head of a youth (cat. 76) used
for Saint Philip in the Last Supper, probably planned at
about the same time as the London Virgin of the Rocks.48 The
dating of these various works, from about 1489 to 1493, is
also consistent with the documents of the initial contract
f i g . 46
l e ona r d o da v i n c i
The Last Supper (detail of fig. 100 showing
Saint Philip’s hands), 1492–7/8
dispute concerning the Virgin of the Rocks, shortly post-dating
December 1490, providing further confirmation of the
starting date of the London work – and thereby clarifying
its relationship to the painting in the Louvre. Interestingly,
the preliminary sketch in the Royal Collection, which is
closest to the initial composition of the National Gallery
picture, also contains architectural elements that are closely
related to features that appear within the unfinished Saint
Jerome; this reinforces the evidence provided by its walnut
support for dating this picture to Leonardo’s Milan years,
started most probably in the very late 1480s.49
c at . 10 (detail)
67
!"#$%&&'()*+,-.*'/012('345(678.9()*+,-.*'%''%&:;:%%''%<9%;''5,=)'%>!
14
15
l e onar do da v i nc i (1452 –1519)
l e onar do da v i nc i
Study of a bear’s head
Studies of a dog’s paw
about 1485
about 1485
Metalpoint on prepared paper
7 × 7 cm
Private collection, New York
Metalpoint on prepared paper
14.1 × 10.7 cm
The National Galleries of Scotland, Edinburgh.
Purchased by the Private Treaty Sale with
the aid of The Art Fund 1991
(d 5189r)
Within the oeuvre of an artist as keenly interested in
nature as Leonardo, his studies of animals assume a
particular value. In addition to his well-known interest
in horses, he made numerous drawings of other
animals, both real and fantastical.
These two sheets contain studies made from life
and share a common provenance from the collections
of Sir Thomas Lawrence (1769–1830) and Captain
Norman R. Colville (1893–1974).1 Both are executed
in the same fine metalpoint technique on light pink
or buΩ prepared paper and date from a period in which
Leonardo was focusing intensely on the animal world.
An analogous work to the Edinburgh study, dated
a few years earlier, is that of the short-haired dog in
the Studies of a dog and cat (British Museum, London,
1895,0915.477). The two dogs probably belong to a
diΩerent breed, as seen from the ruΩled and individual
curled locks of fur on the legs and between the toes
and claws in the Scottish drawing.
Leonardo evinces a similar naturalistic accuracy in
his two drawings of bears – a beautiful head (private
collection) and the Studies for a walking bear and his paw
(Metropolitan Museum of Art, New York, 1975.1.369),
which share similar technique and preparation. Here the
artist demonstrates his profound interest in the beast,
probably an animal in captivity, with a superb economy
of means in the fine, animated metalpoint hatching.2
These drawings have been variously dated between
the end of the 1470s and the mid-1490s.3 Though some
scholars prefer to situate these studies in Florence, they
are better understood stylistically – and hence chronologically – alongside the revolutionary Portrait of Cecilia
Gallerani (cat. 10). Leonardo’s acute observation of
animal anatomy and physiognomy as expressed here
looks ahead to the pictorial invention of the ermine in
that portrait – though with its exaggerated dimensions
note s
1 A common French provenance has
also been suggested: see Weston-Lewis
in Edinburgh, New York and Houston
1999–2000, p. 14.
2 The Studies of a dog’s paw were made from
life from a domesticated dog – and
not a wolf or bear, as has been argued
(Weston-Lewis in Edinburgh, New York
and Houston 1999–2000, p. 14; Barone
and Kemp 2010, p. 46). They have also
been associated with other studies of
dissected bear paws (Windsor rl
12372–5) usually dated to the early
120
!"#$%&&'()*+,-.*'/012('345(678.9()*+,-.*'%''%&:;:%%''%<9%;''5,=)'%>%
literature
Cat. 14: Popham 1937, p. 87; Berenson
1938, vol. 2, p. 115, no. 1044c; Popham
1946, pp. 55, no. 78a, 125; Ames-Lewis
and Wright in Nottingham and London
1983, p. 74, cat. 8; Kemp in London
1989, p. 96, cat. 37; Pedretti 1992,
p. 188; Bambach in New York 2003,
pp. 359–61. cat. 43; Wolk-Simon in
Cremona and New York 2004, p. 89;
Kemp and Barone 2010, p. 106, no. 72.
cat. 10 (detail)
and partially fantastical morphology Gallerani’s
ermine should be seen not as a representation of a real
animal but as a symbolic presence or allegorical figure.4
The combination of fantasy and reality revealed in
the invention of the ermine does not contradict
Leonardo’s conception of the natural, which, as stated
in the Treatise on Painting, can include constructing
imaginary or unknown creatures by assembling their
parts from diΩerent animals.5
Thus both sheets can be seen as important precedents rather than true preparatory studies for the
fascinating creature in the Portrait of Cecilia Gallerani. The
dog’s paws, minutely observed from diΩerent angles,
anticipate those of the wriggling ermine. Similarly,
the ermine’s powerful head may be associated with the
studies of the bear’s, in the structure of the cranium
and shape of its features: small round eyes, cylindrical
muzzle and pointed nose. Furthermore, the present
studies are characterised by their energised luminosity,
with touches of metalpoint evoking the play of light
and shade reminiscent of that on the dense fur of
the animal cradled in Cecilia Gallerani’s arms. ag
1490s (Clark and Pedretti 1968–9,
vol. 1, p. 52; Clayton in London 1996–7,
p. 48; Clayton 2001, pp. 50–1) but
probably from the mid-1480s (I thank
Martin Clayton for this observation).
3 Though frequently dated in the early
1490s, these drawings were correctly
retro-dated to the previous decade
(Popham 1946, p. 55; Weston Lewis in
Edinburgh, New York and Houston
1999–2000, pp. 14–15; Bambach in
New York 2003, p. 357–61; Barone and
Kemp 2011, p. 46), although too rigidly
linked to the end of the Florentine
period. The fact that they are on the
same paper as preparatory studies for the
Adoration of the Magi (fig. 34) and that this
pink prepared paper was not as widespread in Milan as blue prepared paper,
does not appear to be su≈cient
grounds for excluding them from the
Milanese period. Milanese sheets such as
cats 3, 13 and the Studies of a horse (fig. 19)
are similarly coloured and prepared.
4 On the allegorical and symbolic
associations of the ermine see cat. 10.
5 bn 2038 fol. 29r; Urb. fol. 135r;
r 585, m c m 554; k/w 573.
cat. 14
Cat. 15: Popham 1937, p. 87; Berenson
1938, vol. 2, p. 115 n. 1044b; Popham
1946, pp. 55, 125–6, no. 79a/b; Kemp
in London 1989, p. 98, cat. 39; Weston
Lewis in Edinburgh, New York and
Houston 1999, pp. 14–15, cat. 1;
Kemp and Barone 2010, p. 46, no. 4.
cat. 15 (recto)
cat. 15 (verso)
!"#$!%&'()*+,-.*'/012('345(!678.9()*+,-.*':'':";&;::''<=9=>''5,?)'!=!
f i g . 100
l e ona r d o da v i n c i
The Last Supper, 1492–7/8
Tempera and oil on plaster,
460 × 880 cm
Santa Maria delle Grazie, Milan
252
!"#$!%&'()*+,-.*'/012('345(!678.9()*+,-.*':'':";&;::''<=9=>''5,?)'!=>
!"#$%#%&'()*+,-)&./01'&234'!567-8'()*+,-)&9&&9:;<;99&&#=8=:&&4+>(&%##
!"#$%#%&'()*+,-)&./01'&234'!567-8'()*+,-)&9&&9:;<;99&&#=8=:&&4+>(&%#9
91
l e onar do da v i nc i (1452–1519)
Christ as Salvator Mundi
about 1499 onwards
Oil on walnut
65.5 × 45.1 cm
Private collection
It has always seemed likely that Leonardo painted a
picture of Christ as the Saviour of the World.1 In 1650
the celebrated printmaker Wenceslaus Hollar signed
an etching of Christ raising his right hand in blessing,
holding a transparent orb in his left, with a nimbus of
light behind his head; the image was taken, he states,
from a painting by Leonardo (fig. 111).2 Though Hollar
was generally well-informed, this would not be enough
on its own to prove that an autograph picture by
Leonardo had once existed. By the seventeenth century
any number of paintings by his pupils and associates
were firmly attributed to Leonardo himself and there
was no shortage of pupils’ pictures depicting the
Salvator Mundi, all clearly related to one another, all
unmistakably Leonardesque. In 1978 and 1982 one
of these many versions was promoted as Leonardo’s
lost ‘original’, partly because of its similarities to the
etching, a suggestion that has rightly been rejected.3
Hollar might very well have been copying a copy.
There is other evidence, however, that Leonardo
explored this or a related subject. As early as the mid1480s he drew a ‘head of Christ’, in pen and ink, which
appears in the list of his works preserved in the Codex
Atlanticus (see p. 25). And in the early sixteenth
century he discussed painting an adolescent Christ
for Isabella d’Este, Marchioness of Mantua.4 Most
importantly, there survive two red chalk drawings of
draperies, obviously related to the composition etched
by Hollar and the many workshop copies (cats 89, 90).
But even these do not constitute proof that Leonardo
painted a Salvator Mundi, and it has sometimes been
argued that these drawings might have formed the basis
for one or more finished designs – perhaps cartoons –
that he made expressly to be copied by pupils but
with no primary version by the master himself. Other
scholars have imagined, more straightforwardly, that
Leonardo’s own painting disappeared long ago.
The re-emergence of this picture, cleaned and
restored to reveal an autograph work by Leonardo,
therefore comes as an extraordinary surprise. Though
Hollar’s Christ is very slightly stouter and broader, the
two images coincide almost exactly. The draperies are
just a little simplified and there is no glow of light
around Christ’s head. Otherwise the newly discovered
painting has the same snaking locks of hair, expressionless face and uncannily direct gaze, and the same
swathe of monumental drapery across his shoulder.
And the knot-pattern ornament on Christ’s crossed
300
fig. 111
wence slaus hollar (1607–1677)
After Leonardo da Vinci, Salvator Mundi, 1650
Etching, first state, 26.4 × 19.0 cm
The Royal Collection (rl 801855)
stole and on the border of his vestment are very similar
indeed, a particularly important consideration given
that this ornament is the aspect most subject to change
in the diΩerent surviving versions. There can be no
doubt that this is the picture copied by Hollar.
In fact this version of the Salvator Mundi is not a
new discovery. It has been known since the beginning
of the twentieth century but never seriously studied
and certainly not recognised as Leonardo’s own work.
The picture was acquired in 1900 by Sir Francis Cook
for his collection at Doughty House in Richmond,
Surrey, through or from his long-standing adviser, Sir
J.C. Robinson. It has not yet been discovered where
Robinson obtained it. In 1913 Tancred Borenius catalogued it as a ‘free copy after Boltra≈o’, twice removed
therefore from Leonardo. In 1958 it was sold from the
Cook collection, still as a copy after Boltra≈o. The low
esteem in which it was held is easy to explain: by the
time it came into Francis Cook’s possession it had been
very considerably overpainted. Christ’s blessing hand
was the least altered area but his head had been almost
entirely reinvented. And that after 1958 it was known
only from the poor black-and-white photograph
reproduced in Borenius’s catalogue only compounded
the problem.
The reasons for such abundant overpaint are also
clear. Though both Christ’s hands are well preserved,
elsewhere the picture has suΩered. Sometime in the
past the panel split in two, causing paint losses along
the length of the crack. It has also been aggressively
over-cleaned, with some abrasion of the whole picture
surface and especially in the face and hair of Christ,
where Leonardo’s sequence of delicate paint layers
literature
Borenius 1913, p. 123; Suida 1929,
p. 140; Clark 1935, vol. 1, p. 80;
Suida in Los Angeles 1949, pp. 85–6;
Heydenreich 1964, p. 109; Snow-Smith
1982, pp. 11, 12, fig. 7.