A STUDY OF FRENCH PAINTING CANVASESKatrina VANDERLIP CARBONNEL
2 PART I2.1 PREPARING THE SAMPLES2.1.1 CUTTING A SQUARETHE FIRST STEP in preparing the samples for study was to cut a square piece of canvas from the fragment of cloth removed from the painting. When possible, the squares were 10 cm � 10 cm, which was the case only when dealing with the transferred canvases. The square samples cut from tacking margins were sometimes only 1 cm � 1 cm. The samples are made square to facilitate multiplying their surface area to obtain their weight per square meter. Also, the square gives the same accuracy in counting the threads per centimeter in both the directions of the warp and of the weft; for example, threads counted along 10 cm, then divided by 10 give a more accurate count per centimeter than just counting the threads along one centimeter. 2.1.2 CLEANINGTHE NEXT STEP in preparing the samples was to remove the remaining dirt, ground, and paint layers. The sample was soaked in water to swell the fibers and to loosen the particles of ground and paint caught in the interstices of the cloth. The particles were then delicately mechanically removed with a scalpel. 2.1.3 RAVELING THE EDGESTHE SQUARES HAD been cut slightly larger than the intended size of the sample. The edges were then raveled off until the sample was the desired size. The raveling was done after the sample was cleaned so that, should a thread ravel off during the cleaning process, it would not affect the final size of the sample. The samples assumed to be hemp or linen were raveled off while wet as these two fibers are stronger when wet, while the samples assumed to be cotton were unraveled when dry. The stronger the thread being raveled off, the easier it was to remove it. The reason the sample was fringed in this way was to facilitate counting the threads; it is 2.1.4 WEIGHINGTHE FRINGE WAS cut off the sample just before it was weighed. The weight of the sample was multiplied in order to obtain the weight per square meter of the canvas. |
