Rubens and Rembrandt, for example, were equally versatile on
wood and canvas, although they operated within a tradition that had
hitherto favored wood for painting supports. In his Leiden and early
Amsterdam periods, Rembrandt worked almost exclusively on wood, pro-
ducing those beautifully wrought, surprisingly colorful small panels that
established his reputation as a fijnschilder(fine painter). He continued to
use panels throughout his career, but with the production of larger por-
traits and history pieces in the early 1630s, he increasingly chose canvas.
This choice can be easily explained by practical or financial considerations.
Apart from the existing tradition of using panels, they were easily available
ready-made in a range of standard sizes from specialist panel makers. They
were also much preferred for smaller-format pictures because they were
self-supporting and needed only simple preparation. Since panels were
more expensive than canvas, however, there came a point at which it was
worthwhile to go to the greater trouble of stretching and priming canvas.
In his down-to-earth discussion of the advantages of canvas in his Inleyding
tot de Hooge Schoole der Schilderkonst (Introduction to the High School of
Painting),Rembrandt’s pupil Samuel van Hoogstraten wrote that canvas
was “suited most for large paintings and, when well primed, easiest to
transport.”
Incidentally, while on the subject of seventeenth-century Dutch
panels and ofRembrandt in particular, we must note the extraordinary
success ofdendrochronology (up through the seventeenth century) in
clarifying dating problems. Tree ring analysis can also give spectacular
confirmation that certain panels have come from the same tree. For
example, The Woman Taken in Adulteryof 1644 (National Gallery, London)
and the Portrait of Herman Doomerof 1640 (Metropolitan Museum of Art,
New York) have the same structure, as does the panel to which the 1634
canvas painting Saint John the Baptist Preaching(Staatliche Museen zu
Berlin, Preussischer Kulturbesitz, Gemäldegalerie) has been affixed—proof
that the third painting was also done in Rembrandt’s studio. Nor can the
dendrochronologist ignore wider aspects of European history. Peter Klein
recently reminded me ofa long-forgotten war between Sweden and
Poland in the late 1640s, which stopped forever the supply of Baltic oak
to western Europe and established one of those key dates that every
student of painting techniques should know.
Rubens’s restless genius resulted in many extraordinary experi-
mentations with his painting supports. He frequently enlarged his panels
as he went along, in some cases doubling or tripling the original size with
a bewildering patchwork of added pieces. In the Watering Place (National
Gallery, London), composed of eleven pieces of wood, Rubens succes-
sively moved the position of the sun to the left as he extended the compo-
sition. The result is there are now three suns: two painted and one drawn.
In other cases, the complexity of the structure is not the result of enlarge-
ment, since Rubens seems deliberately to have constructed composite pan-
els that were then painted in a single campaign of working. The exquisite
companion paintings of the Château de Steen(National Gallery, London)
and The Rainbow Landscape (Wallace Collection, London), on twenty
planks and nineteen planks, respectively, are examples of this method.
Beyond simple enlargement, did Rubens have a purpose in constructing
such elaborate panels? Did he believe that such construction might some-
how make the work more stable? Was he simply using up scraps of wood?
About half of Rubens’s entire output of oil paintings, including his oilxviii Bomford