The radiography of museum objects is not a new application of the scientific use of x-rays. In fact, in 1896, shortly after Roentgen discovered the x-rays, Doctor Konig radiographed a painting to detect alterations, marking the beginning of an application that is now a routine, valuable part of the total scientific examination of art and archaeological works.
Many people hold the erroneous belief that a radiograph can prove definitely that a work of art is authentic or a forgery. It is not necessarily so. For example, suppose that a forger made a terracotta figurine as an exact replica of an ancient piece and artificially aged it. The radiograph might be unable to show any differences in structure. The final analysis would have to be made by employing other scientific techniques. There is also the misconception that radiography will permit visualization of hidden or covered-over signatures on paintings. Again, not necessarily, because a large percentage of signatures are made with black paint, a carbon derivative that generally will not show on a radiograph.
This is not to parade a series of reasons for not using radiography in the museum: rather, the opposite. So we should mention the positive applications. Radiography is a non-destructive method for visualizing the internal structure of an object. In the art and archaeological world nothing is two-dimensional. Even the thin paper a Cranach engraving has been printed on has internal structure, just as does the 13,35-meter bronze Great Buddha statue at Kamakura. Both have been radiographed for a study of their structure. From this can be learned methods of fabrication, condition, possible restorations and even authentications. Or, in some cases, evidence of fraud.
How radiography can add to our knowledge of museum objects will be explained by means of several examples (because so much has been previously written about the examination of paintings, they will not be included).
The non-destructive aspect of radiography cannot be overemphasized. X-rays used for radiographic examination will not damage an object. Consider the radiograph of a Chinese wooden figure shown in Figure 1. There is a dark, vertical, rectangular image in the body in the area of the hands that indicates a cavity. In this cavity, but seemingly between the hands, we can make out several small white shapes that are metallic packets placed in the cavity, which is sealed at the back. What is actually in the packets will not be known unless the seal is broken and, so far, it hasn’t been. What is important is that radiography has shown what may be votive jewelry sealed in the body. This information could not have been obtained by means of any other technique without destroying the exterior of the statue and without incurring the traditional wrath of the gods.
The use of radiography to reveal repairs is shown in Figures 2 and 3. This graceful Capo di Monte porcelain was on the fireplace mantle in my house when our son was small in size but great in ball-throwing ability. The figurine ended up in a number of pieces on the hearth. Fortunately, a competent restorer was found who was able to put the dancing girl back together so cleverly that the damage was absolutely undetectable to the eye. The girl was actually radiographed so that we could prove to our friends that such a perfect restoration was possible. A study of the radiograph shows the white image of a metal pin in the stem above the base. This stem had been shattered, as is shown by several dark lines across the stem in this area, which are fracture lines. There is an obvious lack of continuity in the area of the chin and cheeks of the face. The irregularly circular region of the chin is new clay that was added to replace a missing piece that we never did find. The dark axial lines in the arms are normal shrinkage cracks and do not show on the surface.
The restoration of the dancing girl was obviously not done with fraudulent intent. But let’s look at a restoration that was so beautifully done that the piece in question was consecutively sold into several large collections as an intact work of art in completely original condition. The terracotta head from Mexico was a fabrication of the Zapotec culture, probably to be used as an insert in a decorative building frieze (Figure 4(. There is an irregularity in the contour of the jaw on the face’s left side, and the clay is quite thin in this area. This is not, in itself, cause for questioning the piece. Nonetheless, the thinned area of the jaw was radiographed in the same manner a dentist x-rays teeth, with a small dental x-ray film packet placed in the mouth (Figure 5). The two dark lines in the radiograph are cracks that have been carefully filled with an adhesive containing some powdered clay taken from the inside of the statue to make the breaks invisible. After a detailed study, the tentative conclusion was formed that a piece of the jaw had broken out due to the lack of strength in the thinned area. When this happened is anyone’s guess. At any rate, the original piece was reinserted in the jaw and the integrity of the surface was restored. The radiograph shows that the texture of the clay in the piece between the two fracture lines is identical to that in the adjacent areas so there is no evidence that a new filler piece was used. Because the head was sold as being in an undamaged condition, fraud was definitely perpetrated.
There is almost no literature dealing with the radiography of textiles, and yet it is a technique that is most valuable in gaining knowledge of weaving methods. Museum curators of textiles are always interested in such details as to how a loom was set up, and the arrangement of the yarns in tapestry as well as in embroidery weaving. Charting these arrangements of yarns is tedious, and even with a binocular microscope, one often has difficulty following the passage of a specific yarn.
Figure 6 shows a photomicrograph of a piece of crepe cloth, and on viewing it we feel that we have a pretty good idea as to how the cloth was woven. Now compare it with Figure 7. This is a radiograph of the same piece of cloth, and immediately an entirely different picture is apparent. We can easily trace the arrangement of each warp and weft of yarn, and see loops that are not evident in the micrograph because they are below the surface. Not being a weaver, I am still somewhat confused at the disparity of the images, but remember that radiography makes a two-dimensional image of a three-dimensional object.
Because of the decorative beauty and intricate techniques encountered in pre-Columbian Peruvian weaving, several of these pieces were used as test objects. Through the kindness of the Brooklyn Museum, representative pieces of embroidery and weaving were made available to us. The first (Figure 8) is of woven natural cotton cloth, 74cm. x 78cm., with four standing figures of wool brocade or embroidery in red, tan, yellow, and brown. This piece was chosen to see if it would be possible to visualize the passage of the individual threads and the stitches employed by the maker, in spite of the fact that they are superimposed on a cloth background. The radiograph of one of the figures (Figure 9) is striking. Not only was it possible to make out the pattern of the cotton cloth, but we were also able to see clearly each thread used to create the figures. Studying the radiographs under magnification gave even more information as we could easily see the double warp yarns (Figure 10).
I have been unable to explain why the color areas radiographed differently in density, as would be expected if metallic dyes were used. According to D’Harcourt, metals might have been employed but in all probability, the colorings were from vegetable or animal sources that should not show up differentially on the radiographs.
The other piece (Figure 11) is a slit tapestry of late pre-Inca period. Of special interest from a technical standpoint are the beaten wefts or horizontal yarns used to produce curved lines. How this was done can be easily seen on the radiograph (Figure 12). Look in the area of the ears, and you can see how the horizontal yarns have actually been pushed or beaten into curves. The black spots and vertical areas indicate the slits. In the slit technique of weaving, two surfaces with a long common border parallel to the warp (vertical) yarns are not joined, resulting in a slit which has an ornamental function.
Not all types of textiles are suitable subjects, as heavy multiple yarn weaving or embroidery will be too complicated radiographically. Interpreting the image would be somewhat comparable to tracing one long piece of spaghetti in a bowl full.
The final example of textiles was found fairly recently in western New York, south of Rochester, An Indian burial site was excavated, and rolled up under the skull was a uniform-like jacket that we were unable to open up because of its fragile condition. Certain areas of the cloth bundle were of a malachitic green color and heavily corroded, indicating the presence of metal (Figure 13), A small piece of this cloth was carefully removed and radiographed (Figure 14). Each yarn was wrapped diagonally with thin metal foil, later identified as copper, and possibly originally coated with a silver or gold wash, forming what might be referred to as gold braid. We certainly could not see this type of information as easily with the microscope.
The next example shows how the radiography of an object can result in a frustrating but humorous chain of events and end up without any positive answers. A number of years ago 1 became involved in a detailed radiographic study of paper. In an attempt to work with all forms of paper a museum was contacted concerning the loan of some fragments of papyrus from Egypt, and we soon were working with several dozen specimens. To our amazement, the radiograph of one piece was unlike all the others. The image of the papyrus contained many minute specks, such as we might expect from finely ground metallic powder (Figure 15). Microanalysis showed that these minute particles were sodium chloride, or common salt. They were not on the surface of the piece or between the three layers of papyrus that make up the sheet, discounting the notion that the piece had been soaked in salt water after it had been made. Instead, they were within the very fibers of the plant material from which the paper was made.
The first thought was that the plant had grown in a briny environment, but museum authorities repudiated the idea that the salt might have been drawn up into the plant through the root system. The scientific approach would be to obtain a living papyrus plant, feed it brine and x-ray the stalks. So we finally located one, and had it shipped to Rochester. The potted plant turned out to be higher than the ceiling in my office and we had to remove some of the ceiling panels before we could stand it upright. The plant died almost immediately, in spite of the ministrations of a myriad of ladies with “green thumbs,” and before we could feed it salt water. So radiography can lead to unanswered frustrations.
Strong positive answers can often be obtained through radiography in the examination of metal objects, especially if they are heavily corroded. The example chosen reveals how radiography can help in proving a piece lacks integrity through a study of its structure. The example is of double interest in that it is practically identical to one in the collection of the University Museum,
The object in question is a 29-inch long bronze sword from Iran (Figure 16). It is heavily corroded, and from the standpoint of the patina seems to be normal. The radiograph and drawing (Figure 17) caused consternation because I had just purchased the piece for my collection. In the vernacular, I had been “taken!” Normally the blade would have had a long tang that went up into the hilt. The blade actually has no tang, and barely fits between the two wings of the pommel. The white area in the radiograph indicates modern solder, chemically treated to appear corroded.
I wrote my findings to Maude de Schauensee of the University Museum staff and received another jolt—the Museum study collection contained a sword that was so similar in construction that both could well have been fabricated by the same person. Of interest is the fact that both the blade and the hilt are authentic, but never were made for each other. As Gayle Wever, also of the Museum staff, explained, the hilt originally held an iron blade, the broken tang of which is still in the hilt. Someone took the two old pieces and joined them together, thus fabricating a forgery. Probably, I would still be content if I had never radiographed the piece.
Suppose that a polychromed Japanese statue is brought to you with the request to find out whether the body was carved from a single piece of wood or was made up of several pieces fastened together. Visual inspection is negative. A radiograph made in the front-to-back direction is not conclusive (Figure 18). Some wood graining is apparent but difficult to follow throughout the body area. The white circles around the head are brass ornamentation. A radiograph is then made of the side view of the statue, a magnified section of which is seen in Figure 19. The faint vertical parallel lines are the grain of the wood and are continuous through the head and body, showing that basically it was made from one piece of wood. The exception is the vertical white line running through the face just in front of the ears. This is a glue joint. The face was made separately, so that the eyes could be inserted from the back into their sockets. Then the face was glued into place and the joints smoothed until the surface was continuous. The polychrome easily prevented the joint from being visible.
It would be impossible to describe all of the interesting applications of radiography to art and archaeological objects in an article this brief. We have shown, however, that the radiographic study of internal structures can aid greatly in the determination of integrity within a piece. Restorations for preservation or with fraudulent intent are discernible as well as information concerning fabrication. As new radiographic techniques are developed, we will undoubtedly be able to apply them to an even greater spectrum of objects.