In 1906, on the occasion of the Rembrandt jubilee, Jozef Israels bore witness to his lifelong admiration of Rembrandt and his art, conjuring up a picture of the master working on the Night Watch. The vision he evoked was of a painter in the throes of creation, 'dipping his broadest brushes deep into the paint of his large palette' in order to give more power and relief to certain areas of the painting. The author contends that this description is not consistent with what really went on in 17th-century studios. Numerous arguments support the hypothesis that up into the 19th century palettes were not only much smaller than the 19th-century ones envisioned by Jozef Israels, but that they did not usually carry the complete range of available oil-based pigments. On thc contrary, painters adhered to the diehard tradition of loading their palettes with a limited number of tints suitable for painting a certain passage. Support for this proposition comes from various directions. The most important sources are paintings of studio scenes and self-portraits of painters with their palettes. Examination of the depicted palettes, an examination conducted on the actual paintings, has yielded plausible grounds for assuming that painters strove for verisimilitude in their renderings of palettes. This is borne out by the surprising consistency of the examined material. On certain 15 th and 16th-century representations of St. Luke painting the Madonna, his palette is seen to contain only a few shades of blue, with occasionally white and black. Other palettes on which a greater variety of colours are depicted are incomplete, representing the range needed for the parts of the painting which were the most important and most diflicult to paint - the human skin. Texts by De Mayerne and Beurs gave rise to this assumption. One of the chief duties of the apprentice was to prepare his master's palettes. According to a dialogue in the late 17th-century Volpato manuscript, the master's mere indication of which part of the painting he was going to work on sufficed for the apprentice to prepare the palette. This implies that a specific number of pigments were necessary for the depiction of a particular element of reality. The idea is supported by the countless recipes for the depiction of every part of the visible world which have been handed down to us, notably in Willem Beurs' book but in other sources too. The implication is that the method of a 17th-century artist differed fundamentally from that of artists of the second half of the 19th century and the 20th century. Whereas there are substantial grounds for assuming that painters of the latter period tended to work up an entire painting more or less evenly, painters of earlier centuries executed their work - over an underdrawing or an underpainting in sections, on a manner which is best compared with the 'giornate' in fresco painting. This kind of method does not necessarily mean that a painter did not proceed from a tonal conception of an entire painting. Indeed, Rembrandt's manner of underpainting shows that his aims did not differ all that much from, say, Jozef Israels. Technical and economic circumstances are more likely the reason why painters continued to work in sections in the Baroque. With regard to the economic aspect: grinding pigments was a lengthy operation and the resulting paint dried fast. Consequently, no more pigments were prepared than necessary, so as to avoid waste. With regard to the technical aspect: before the development of compatible tube paints, whose uniformity of substance and behaviour are guaranteed by all manner of means, painters had to take into account the fact that every pigment had its own characteristics and properties; some pigments were not amenable to mixing, others were transparent by nature, other opaque, etc. This is best illustrated by paintings of the 15th and 16th centuries. However, the tradition persisted into the 17th century and was also carried on by Rembrandt, as scientific research has shown. Neutron-activating radiographic examination reveals that certain pigments only occur in isolated areas (as far as these pigments were not used in the monochrome undcrpainting). Scrutiny of paint samples has moreover revealed that a layer of paint does not as a rule contain more than two to five, or in very exceptional cases six, pigments. Having been made aware of this procedure, however, we can also observe it in stylistic characteristics of the painting, and we realize that for the aforesaid reasons a late Rembrandt is more akin to a Raphael than to a Jozef Israels. In the 19th-century discussion of the relationship of style and technique, figures like Semper contended that this relationship was an extremely close one. Riegl, proceeding from the concept of 'Kunstwollen', regarded technique as far less important, more as the 'frictional coefficient' in the realization of a style; while not denying technique's effect on style, Riegl did not consider its influence to be as crucial as Semper did. Paul Taylor's recent research into the concept of 'Houditng' have demonstrated the extent to which aspects as tone and colour served to create an illusion of space in the 17th century, the chief priority being the painting as a tonal and colouristic entity. If we assume that the working principles of a 15th and a 17th-century painter did not fundamentally differ, it becomes clear that the pictorial 'management' involved in attuning tones and colours so convincingly as to produce the tonal unity so typical of Baroque painting, was quite an achievement. The technical and economic limitations mentioned above in connection with the palette may thus be seen as exemplifying Riegl's view of technique as a frictional coefficient in achieving pictorial ends.