JAIC 1990, Volume 29, Number 2, Article 2 (pp. 117 to 131)
JAIC online
Journal of the American Institute for Conservation
JAIC 1990, Volume 29, Number 2, Article 2 (pp. 117 to 131)





ERASERS WERE tested to determine whether oil, plastic, or rubber components traditionally present had any adverse effects on the chemical watermark. The erasers chosen were Artgum, Kneaded, Magic Rub, Opaline, Pink Pearl, Staedtler Mars Plastic, and Groomstick. The watermarked area of the paper was erased 25 times in a small circular motion characteristic of traditional cleaning methods. Any changes were noted.

The crumbs from the erasure procedure were gathered and laid on the watermarked area. A thin polyester film was placed over the crumbs to reduce displacement possibilities. The papers were then stored on top of a blotter in a metal flat-file drawer away from light. After eight weeks, the papers were examined, and changes were recorded.

7.1.1 Results

The plasticizers, oils, or rubber components in the erasers used for this test did not appear to alter the watermarks. With the exception of Groomstick, the erasers did not seem to damage the paper. As viewed under normal illumination after eight weeks, all three papers darkened in the area in contact with the Groomstick. The darkened area in the JCC sample fluoresced a dull, opaque white under both short- and long-wave ultraviolet illumination.


A variety of organic solvents is used to remove stains and adhesives from paper. To determine the parameters of susceptibility to damage from such applications, the following solvents were chosen for testing: acetone, amyl acetate, carbon tetrachloride, diacetone alcohol, dichloromethane, dimethylformamide, ethanol, methanol, methoxy magnesium methyl carbonate in methanol and trichlorotrifluoroethane, methyl ethyl ketone, naphtha, petroleum benzine, 2-propanol, tetrahydrofuran, trichloroethylene, toluene, and xylenes. Acetic acid (2% in deionized water, pH 5.5) was also used. Three watermarked areas of each sheet were tested as follows: one drop in each area, allowed to evaporate thoroughly, followed by another drop in two of the areas, thoroughly dried, and ending with a third drop in one of the areas. A nonwatermarked area was similarly exposed to each solvent for comparison. Testing was completed within the fume extraction hood, and observations were noted.

7.2.1 Results

The chemical watermarks were moved or eliminated with the application of every solvent chosen. A ring of what appears to be watermark material was found encircling the test spots in all papers after the solvents evaporated.

Under short-wave ultraviolet illumination, the watermarks appear unaltered, with well-defined edges. The exceptions were the watermarks tested with acetic acid and dimethylformamide. The application of acetic acid created lighter white spots; the application of dimethylformamide created dark bluish dots encircled by a very bright fluorescent ring formed by the movement of the optical brightener in the paper.


The water-based solutions sat on the surface of the papers, making the dropper technique an inefficient test method. Therefore the papers were immersed in trays of the aqueous alkaline solutions.

The solutions used for testing were deionized water (pH 7.0) ammonium hydroxide in deionized water (pH 8.0–9.0); calcium hydroxide in deionized water (pH 8.0–9.0) and a saturated solution of calcium hydroxide in deionized water (pH 12.0). For each solution samples of all three papers were immersed in the same bath, and polyester film was placed on the water surface to ensure immersion of the papers and to prevent evaporation of the liquid. One sheet of each sample was removed after 15 minutes; the remaining three papers were removed after 2 hours. Upon removal, the samples were left to air dry on a polyester screen. After drying, they were compared to a control piece in normal and normal transmitted illumination and under ultraviolet sources. Observations were recorded.

7.3.1 Results

Traditional watermarks become more translucent when wet. The chemical watermarks also reflected the color of the underlying trays to an unusual degree. While drying, the watermarks appeared to be opaque and diffusing into the paper. Once completely dry, the watermarks again became translucent and the papers returned to being opaque.

After drying, there was some blurring of the watermarks in most of the papers. This loss of sharpness seemed to increase with longer immersion times. It was particularly apparent in the GE shadowmark sample and least apparent in the nonbrightened JCC sample.

Under short-wave ultraviolet illumination, the watermarks appeared faint and indistinct from the paper. A fluorescing component or components from the watermark or paper may have been solubilized and redeposited across the surface, shielding the watermark from the ultraviolet source.


Seven chemicals used for bleaching and bleach neutralizing (antichlors) were tested on the watermarks and the surrounding paper areas: chloramine-T (sodium toluene-4-sulphon-chloroamide), chlorine dioxide, hydrogen peroxide, sodium borohydride, sodium hypochlorite, sodium meta-bisulfite, and sodium thiosulfate. A 5% solution of each chemical in water was prepared for testing. (The exception was chlorine dioxide, which was prepared at 2%.) Solutions at a concentration of 5% are higher than those normally used in conservation practice and were chosen as extreme conditions under which to test the watermark material. The concentrations were kept the same so that the effects could be compared more accurately.

Most of the testing was done in a fume extraction hood. In each case, the chemicals were poured onto the papers in the area of the watermark and allowed to sit for 15 minutes. The excess was drained off, and the papers were blotted, then allowed to air dry on a polyester screen. Afterward, the papers were immersed in a deionized water bath for one-half hour and again allowed to air dry. The papers were observed with normal, normal transmitted, and long- and short-wave ultraviolet illumination. The effects of the chemicals on the watermarked designs were recorded during the procedure, after the bleaching, and after the final rinsing.

7.4.1 Results

Testing produced inconsistent and contradictory results. The type of watermark, whether line or shadow, seemed to determine whether a bleach or bleach neutralizer was detrimental. Further testing with various concentrations and complete bleaching procedures might help obtain more applicable information.

Chloramine-T diminished all three watermarks. This effect was noticed visually with normal transmitted light and under long- and short-wave ultraviolet illumination. Sodium thiosulfate and sodium meta-bisulfite either blurred or obliterated the BG and GE watermarks; the JCC paper had a mottled fluorescence across the surface, but the watermark remained sharp.

Hydrogen peroxide left the BG watermark fainter under all types of illumination but did not seem to damage the GE and JCC watermarks. Watermarks treated with sodium borohydride appeared very faint but retained sharpness of definition under all illumination methods. Sodium borohydride also caused the papers to blister. Chlorine dioxide and sodium hypochlorite did not appear to alter any of the watermarks. Watermarks tested with both of these bleaches appeared dark on the obverse and fluoresced on the reverse under short-wave ultraviolet illumination. The appearance of the watermark under ultraviolet illumination before testing was dark on the reverse and unseen or white on the obverse.

All the chemicals except hydrogen peroxide and sodium thiosulfate eliminated the cream colorant in the JCC paper.


Naturally aged, chemical watermarks in the collection of the Dard Hunter Paper Museum at the Institute of Paper Chemistry (now the Institute of Paper Science and Technology, Inc.), seem to be diffusing into the paper support and disappearing. Yet when viewed under short-wave ultraviolet radiation, the designs look crisp and precise. Natural aging, cross-linking of the material, or uncontrolled fluctuating environmental conditions in the museum may be the cause of this phenomenon.

In an attempt to simulate the diffusion of the watermarks, the papers were subjected to two types of artificial aging. Three samples (one sample sheet of each paper) were placed in the RH-controlled environmental aging chamber (Blue M Electric Humid-Flow Combination Temperature and Humidity Chamber) at 88C ( 0.5C) (190F [ 1F]) at 55%–60% RH for six days. Three more samples were placed in a dry oven (Lab-Line Imperial II Radiant Heat Oven) without exposure to humidity at 100C (212F) for six days.

The papers placed in the humid aging oven darkened and turned brown. The watermarks were darker than the paper and opaque. All translucency of the watermark was lost. The samples placed in the dry oven were slightly brown and, similar to the museum's specimens, had diffuse, translucent watermarks that were still sharp under short-wave illumination.

Copyright 1990 American Institute for Conservation of Historic and Artistic Works