RKD STUDIES

Cusping can sometimes be difficult to observe from the X-radiographs. Consider the bottom portion of the X-radiograph of The Little Street (L11) in figure [1].¹ Angular distortions in typically straight threads due to cusping on the right, bottom, and left edges are highlighted with a trailof red dots. The horizontal threads along the right and bottom edges exhibit only tiny fluctuations.

Compare this to the weave angle maps of The Little Street (L11) in the lower half of figure [2]. The cusping on the top and left edges of The Little Street (L11) is vividly illustrated in the lower images in figure 2. The angle map coloration enhances perception of the presence, depth, and spacing of cusping. The shallower cusping on the bottom and right edges of The Little Street (L11) is very faintly indicated in the angle maps in figure 2.

Why might The Little Street (L11) have strong cusping on only two edges? One possibility is that the canvas was primed while mounted on a large strainer.² The application of sizing or ground would stiffen the pattern imposed by the large cusps from stretching on an oversized strainer. A rectangle cut from a corner of the large primed canvas would have strong cusping on two adjacent sides, as does The Little Street (L11) on its top and left edges as shown in figure 2. The Little Street (L11) is one painting considered to be primed on one strainer, cut down, and mounted on another strainer for painting.³ Faint, so-called secondary cusping, would be expected once the smaller primed rectangle was stretched onto a smaller strainer in preparation for painting. Importantly, The Little Street (L11) retains its primed, but unpainted tacking edges. This confirms that the canvas has retained its original size and was cut from a larger preprimed canvas.

Thus, weave angle maps can be a significant aid in understanding the extent and shape of cusping. They can also reveal information on cusping being primary or secondary. But, often more information is needed to confirm this assessment, such as that provided by the tacking edges, when available, as in the preceding analysis of The Little Street (L11).

A similar history of priming at a large size (thereby inducing primary cusping), cutting out a smaller rectangle, and re-mounting the smaller rectangle for painting (potentially producing secondary cusping) could be associated with a canvas having strong cusping on only one side and secondary cuspingon all others. This could occur with a canvas extracted not from a corner but from the edge of the larger primed canvas. The deep, widely spaced cusping in the horizontal threads across the top edge of A Lady Writing (L20) and the shallower, more closely spaced cusping along the other three edges, makes A Lady Writing (L20) a prime candidate for having been prepared on an oversized strainer.

1
Bottom Portion of X-Radiograph of The Little Street (with red dots tracing threads on right and left edges)

2
Weave Maps of The Little Street (from 0.5 cm evaluation squares centered on 0.25 cm grid)

Another factor in interpreting the extent of cusping from a weave map constructed from an X-radiograph is the assumption that the X-radiograph covers the full painting, that is the X-radiograph has not been trimmed and the X-radiographs made provide full coverage. This need not always be the case. For example, when the film is inserted between the stretcher bars and the canvas, the edges of the painting are not always captured in the X-radiograph. The method of taking X-radiographs with several overlapping X-ray films exposed separately and covering the full painting results in X-radiographs and the painting being to the same scale. Thus, the size of the stitched X-radiographs should match (or possibly slightly exceed) the size of the painting. In some cases the X-radiographs are a bit larger than the painting, for example Woman with a Pearl Necklace (L18), and in others a bit smaller, for example Woman with a Lute (L14). Though the paintings Woman with a Lute (L14) and Woman with a Pearl Necklace (L18) are the same size, their stitched X-radiographs and, therefore, their weave maps are not, as is apparent in figure 7 in § 6.1. To investigate this size relationship, table 1 was assembled.⁴

The largest discrepancy with the X-radiograph smaller than the painting is with The Procuress (L03). Comparing the stitched X-radiograph with the painting it is apparent that the stitched X-radiograph does not capture the full painting. The absence of stretcher bars in the stitched X-radiograph of The Procuress (L03) suggests that the films were inserted between the stretcher bars and the canvas and did not capture the edge of the canvas.⁵ These differences need to be carefully assessed if the depth of cusping is to be determined precisely.

Any trimming of the canvas since the setting of primary cusping at its priming will have an effect on the visible depth of cusping. For example, the absence of cusping along the right edge of Diana and Her Nymphs (L01) is consistent with the observation that Diana and Her Nymphs (L01) has been cutdown after completion with a vertical strip approximately 10 cm removed from the right edge.⁶

Table 1
Dimensions of Stitched X-Radiographs and Paintings

Over half of Vermeer’s paintings exhibit cusping exceeding 5 cm on all four sides.⁷ One scenario for Vermeer’s studio practice that is consistent with this observation is that Vermeer often painted on canvases that were either (i) mounted, primed and painted on the same strainer or (ii) mounted already primed, potentially inducing secondary cusping, and painted.⁸ In scenario (i) the tacking edges would have no primer or paint. In scenario (ii), the tacking edges would be primed but not painted. The weave angle maps for The Geographer (L27) show distinct and evenly distributed cusping on all sides. All tacking edges of this painting have been preserved and are unprimed showing that the canvas was prepared in its present size, that is an example of category (i). Furthermore, the unprimed tacking edges of The Geographer (L27) indicate that The Geographer (L27) was not primed while still connected to its rollmate The Astronomer (L28). The weave angle maps of Young Woman Seated at a Virginal (L34) show only slight cusping on all four sides. The ground – but not the paint – extends onto all tacking edges, which implies that this canvas was cut from the center of a larger primed canvas, that is an example of category (ii). These examples suggest that the depth of cusping, which can be readily accessed from the scaled allVermeerAngleMaps, can be an indicator of primary or secondary cusping when combined with information gleaned from the tackingedges.

Currently, we know very little about Vermeer’s studio practice. The kinds of information that can be extracted from observations of cusping provide some evidence of this practice.