| Abstract: | SUMMARIES. The original layers of paint on Swedish and Finnish furniture and interior decoration of the late eighteenth and early nineteenth century were studied so that their original character could be preserved during restoration. A cyclotron was used for particle-induced X-ray emission (PIXE) analysis. The common light layer on furniture of this time was found to be a lead-based paint applied on the plain wood. In the early nineteenth century zinc white was gradually introduced. PIXE analysis proved to be an outstanding method for evaluating relative concentrations of contaminants. The use of comparative indices instead oj absolute concentrations appears to be justified for cost-benefit reasons. The application of the method to four chairs, a Mora grandfather clock and a door moulding is presented.
CONCLUSION. White and light grey became popular colours in the northern countries with the introduction of the neoclassical style, presumably because it resembled the marble of the buildings of antiquity. The use of light colours may also have been a way ot countering the darkness ot the dull winters at high latitudes. Ever since, white furniture and interiors have been common in these countries. The predilection for white without any tint ot grey or other hues is common in Sweden and parts of Finland. This is exemplified by illustrations from reference 19. pp. 70, 71, 91-98. The excessive use of white has been criticized by the artist 1). Batchelor: 'It did its work on everything around it, and nothing escaped' [13, p. 335]. The use of white lead-based pigments in the eighteenth century and their gradual replacement by zinc white in the nineteenth century is well demonstrated in the present study. No underlying layer of chalk as discussed in reference 20 was found. It is difficult to determine with any certainty the time that zinc white made its breakthrough. The authors suggest that it may have happened earlier than mentioned in the literature, due to its earlier production in the processing of copper. Usually the furniture has been painted many times. After the first layer of lead-based paint, many layers of predominantly zinc white follow and finally, in the last layers, titanium white. Typically, one or two non-fluorescent layers are followed by up to three strongly fluorescent zinc white layers and finally one non-fluorescent layer of titanium white. Tinting with other pigments is rare or difficult to detect by eye in these later layers. White furniture was rare between 1798 and 1827 in Turku. Instead, grey was frequent [4], indicating the custom of tinting the white with some black pigment at that time. The use of some greyish lead-based pigment in painting the furniture is another explanation for the colour. The general practice of repainting furniture several times is disastrous for ornamentation, which becomes obscured by the multiple layers. To remove added layers and preserve the original bottom layer is a tedious task during restoration. In the present study, the use of PIXE to assess the cultural and historical information of paint layers during conservation and restoration work has proved reliable, convenient and surprisingly informative. The method opens up the possibility ot benefiting from concentrations ot elements in the paint within a dynamic range of sometimes tour to five orders of magnitude. Thus it is possible to draw conclusions about the composition and origin of the paint not only from the concentrations of its main constituents but from trace amounts of elements as well. A condition for the present analysis is the large difference in concentrations of elements in the paint of the artifacts to be restored. It allowed us to assume a rough proportionality between the concentration in weight percent, u, and the output values of the GUPIX program, c For unknown thickness, elemental and mass distribution, and composition of the sample, the constant of proportionality K may vary greatly, making the indices i-alone useful. The ratios of comparative indices, say ot zinc and lead, may be useful when comparing different samples. By the use of comparative indices, short exposures to the cyclotron are sufficient for an informative analysis. As a rule. 10 minutes may be enough. |
|---|
