JAIC 1983, Volume 22, Number 2, Article 1 (pp. 57 to 61)
JAIC online
Journal of the American Institute for Conservation
JAIC 1983, Volume 22, Number 2, Article 1 (pp. 57 to 61)


J.K Hutchins

ABSTRACT—This report describes published research that analyzes and explains phenomena familiar to those who work with water-stained cellulosic fabrics and paper, and that suggests a new line of inquiry in the study of foxing.


MOST TEXTILE AND PAPER CONSERVATORS have been confronted with cellulosic artifacts bearing waterstains. The marks are characterized by a brown boundary that varies in intensity, width, brittleness, and permanence. The color of the line ranges from a light, barely perceptible tan to coffee-colored brown. On an undyed fabric the wetted and dried area will usually differ in color from the unwetted area, and as it is often lighter than the brown boundary itself, the brown line is ascribed to dirt and degradation products that are carried and deposited by the spreading liquid. The success of efforts to remove this staining with water or with water and surfactant is apparently related to the age and intensity of the mark, as well as to the initial processing of the substrate. Occasionally the browned margin is severely embrittled, breaking apart if flexed.

Since the fabrics or papers in question are usually old and often soiled, it is reasonable to suppose that the brown lines on such materials are a consequence of the soil and cellulose decomposition products moving with the water by capillary action. After immersing the objects in water, some conservators endeavor to prevent the browning at the edges or thicker, slower-to-dry portions by accelerating drying or by covering the object or its edges with cotton sheeting, net, or paper.3 Browning then occurs on the surface or edges of these secondary materials, which can be removed after drying. If damp cellulosic fabrics such as wet sheets are not dried promptly, a similar browning occurs on the surfaces in contact with the air. However, the occurrence of brown lines on clean cellulosic materials raises the question of how this effect can be due solely to soils. Consideration of the behavior of clean materials suggests that stains observed by conservators may have multiple causes.


In 1934, W.T. Bone published a short communication in which he observed that when an end of a strip of pure bleached cotton fabric was dipped into pure water, a brown line appeared, and that methylene blue, cuprammonium fluidity, and Harrison's solution indicated that the cotton in the browned area at the wet-dry cellulose interface suffered oxidative degradation.4 Only the water-air-cellulose interface suffered damage: the completely wet and the completely dry portions showed no change. The importance of the phenomenon to the dye industry was emphasized in the subsequent discussion, when several speakers commented that bleached cotton or rayon goods left damp would dye unevenly.

A more detailed study was published in 1950 by Bone and Turner5 who suggested that the brown line effect might have five possible causes:

(1) a restricted dispersion of free or slightly aggregated cellulose molecules by water, and their transfer by convection to the boundary region; (2) modification of cellulose over the whole area of the test strip, and the collection of the modification product occurring in the wet part by convection at the boundary line; (3) modification by water or by water and air, in the wet region only, and transport of the modified product to the boundary line; (4) initial presence, or formation during the experiment, in the wet area, of a substance capable of reacting with the cellulose, and the concentration of much of this substance at the boundary line before it has been able to react with the cellulose in the wet region generally; (5) a specific change in the cellulose which can take place in the conditions which exist in the boundary region, and nowhere else. [p. 326].

During an extended series of experiments it was shown that although the brown lines are water- and ethanol-soluble, a fluorescing modification remained at the water line whenever the rinsing was not effected promptly. The authors concluded that, “The presence of a non-dispersible and non-transportable modification at every position in the cloth where a boundary between wet and dry has existed for any length of time reduces the probability of alternatives (1), (2), and (3).” The fourth solution resembles that advanced by conservators hypothesizing motion towards and concentration at the stain edge of existing soil and degradation products.

In the published experiments, the fabric used was a new bleached, scoured, and washed muslin. The authors were careful to avoid contamination of solvent, fabric, glassware, and atmosphere, but found that the brown line still occurred. Repeated use of the same sample caused repeated formation of the line, although a slight diminution of the browning on the second trial was noted. Subsequent lines on the sample did not vary in intensity if similar conditions were maintained. Eliminating impurity as a primary cause of the phenomenon, experiments suggested that the interaction of air, water, and cellulose is responsible for the formation of the brown line. Furthermore, it was concluded that the energy of the water molecules at the interface, while sufficient to permit evaporation, was probably not enough to cause degradation. The authors also demonstrated that the reaction occurs in the absence of light, heat, atmospheric oxygen, iron, bacteria, and waxy materials.

One of the numerous studies of weathering—the effect of light, moisture, and atmosphere—prompted further examination of the results described above. Testing the effects on cotton fabric of repeated wetting and drying, Bogaty et al2 found that the brown of weathered cotton fabric was comparable to the fluorescing material generated in the brown line experiments. Elaborating some of Bone's studies, they were able to confirm several of his findings and to furnish new data.

Like their predecessors, they concluded that the brown line effect did not result from iron or prior existence of the brown substance, and noted that the rate of production was nearly constant in repeated trials over eight months. The degraded, acidic nature of the boundary area was confirmed, and the product subjected to further analysis. The soluble brown was determined to be of low molecular weight, and the breaking strength of the fabric (previously aged at elevated temperatures) was shown to have diminished. When Aspergillus niger was introduced on the fabric sample, initial growth occurred rapidly and exclusively along the brown line, spreading only later to other areas.

Both groups considered their conclusions speculative, and refrained from postulating a reaction mechanism. Such caution is understandable, given the difficulty of identifying “pure” cellulose, excluding all oxygen, controlling the breadth of the line formed, and producing sufficient brown material for experimentation. Schaffer et al.16 were among those assessing the relative importance of atmosphere, substrate, and solvent in the reaction. They concluded that continuous evaporation at a wet-dry interface resulted in a fluorescing brown line (sic), accompanied by modification of the substrate or the solvent. Related studies were performed by Madaras and Turner,11 who examined the evaporation of water from cotton fabric in nitrogen and in vacuo. The latter procedure hindered brown line formation without preventing modification of the substrate at the interface, leading the authors to suggest the formation of – COOH groups despite the absence of atmospheric oxygen. They summarized their agreement with Bone and Turner and with Bogaty and coworkers as follows:

  1. Evaporation of distilled water from a fabric, under the conditions defined, resulted in a sharp boundary between wet and dry regions.
  2. Modification of the cellulose took place at this boundary with the production of a brown, water-transportable, fluorescent reaction product.
  3. After extraction of this brown material, there was evidence of modification of the residual cellulose at the place where it had been formed. The methylene blue absorbtion was markedly greater, and there was evidence of a slight local increase in cuprammonium fluidity as compared with that of the wet or dry regions of the cloth.
  4. Periodic lowering of the evaporation level in the same cloth during the same experiment gave rise to a series of fresh brown lines of undiminished intensity, although, at each lowering, the area of wet cloth from which water could evaporate was correspondingly reduced.

Observation (4) seems to be conclusive evidence that the brown line does not represent merely a concentration at the boundary of non-cellulosic material present originally in the cloth sample or formed during the experiment in the whole of the wet region. [My emphasis]

The authors also note that if oxygen is required for brown line formation, the amounts are “minute” (p. 374).


THESE FINDINGS offer sound and provocative indications that the often-observed water-marks are sources as well as consequences of damage to cellulose: browning is not simply unsightly but evidences oxidation. In the treatment of aged or soiled cellulosic materials, the presence of degradation products and soil must not be discounted, but it is equally clear that they are not solely responsible for the observed staining and embrittlement of areas once wetted. Furthermore, improper drying will cause damage to cellulosic materials. The browning of the last-to-dry page edges or trapunto design is a problem that is both cosmetic and chemical. The various reactions involved—the degradation of cellulose, the interaction of water and cellulose, and the chemistry of browning—have been considered separately by various authors.7, 8, 9, 10, 13, 15, 17


BESIDES EXPLAINING the consequences of spotting cellulose with water, this literature could offer a foundation for the study of the problem of foxing. Speculation and published data provoke the question of whether certain types of the reported spotting and staining known as foxing are related in any way to the brown line reactions.

There has been extensive discussion of the brown spots occurring on paper, with most authorities attributing them either to iron accumulated in the paper during processing, or to fungal infection. Recent authors1, 12, 14 seem to favor the latter explanation.

Using fluorescence microscopy, Meynell and Newsam examined samples from eleven books and, although the books were not obviously mouldy, found some hyphae associated with the foxed areas. Like the brown in the experiments discussed above, such spots were acidic and fluorescent. While suggesting that the growth of mould was responsible for the reaction of the amino acid indicators at the fluorescing perimeter of the stains, the authors noted that growth occurred around the lesions, that it was remarkably sparse and slow, and that it had no apparent effect on the fibers or on the paper. They concluded that growth occurred on the size, or on the soiled page edges; and that it was slowed by environmental fluctuations or insufficient food. Rather surprisingly, they then suggested that the tissue paper beside an engraving foxes readily precisely because it is unsized, and tends to be damp and absorbent. The role of sizing on moisture sorbtion is not considered, nor is the possibility of a correlation between browning and unevenness of size application.

If, in damp conditions, the page edges become the air-cellulose-water interface, brown line studies would predict local damage. Spotting on the rest of the page might be a function of variation in sheet thickness, leading to uneven concentration of moisture and, in areas slowest to dry, browning. As Bogaty observed, mould grew preferentially in the brown line-damaged areas, suggesting that the observed growth might be a result of spotting rather than a cause.

Such a view is consistent with the findings of Cain and Miller,6 who distinguished five different classes of foxing. Two of the classes were studied, and one was persuasively attributed to the presence of iron particles. In the group designated class 2, it was found that the local iron content of one sample group was lower than that of the control, and that in another sample group, hyphae were associated with only one of three hundred sixty spots.

Given that class 2 spots are not yet firmly attributed to iron and that fungal activity is not solely responsible for spotting, the effect of fluctuating moisture content on a cellulose substrate of uneven thickness should be considered.


THANKS are due to Dennis Piechota of Arlington, MA for his comments; and to Helen Burgess of the Canadian Conservation Institute for suggesting additional references.


Baynes-Cope, D., “Some observations on foxing at the British Museum Research Laboratory,” Int. Biodetn. Bull. 12 (1976): 31–33.

Bogaty, H., Campbell, K.S., and Appel, W.D., “Some observations on the evaporation of water from cellulose,” TRJ22 (1952): 75–82.

Bogle, M., “The uses for organic solvents in textile conservation,” ICOM 6th Triennial Meeting, Ottawa1981.

Bone, W.H., “Evaporation of water from cellulose,” J. Soc. Dyers and Colourists50 (1934): 307–309.

Bone, W.H. and Turner, H.A., “Some effects of the evaporation of water from cellulose,” J. Soc. Dyers and Colourists66 (1950): 315–327.

Cain, C.E., and Miller, B.A., “Photographic, spectral and chromatographic searches into the nature of foxing,” Preprints of the Tenth Annual Meeting of the AIC, Milwaukee1982, 54–62.

Davidson, G.F., and Standing, H.A., “Auto-hydrolysis of acidic oxycelluloses,” TIJ Transactions (1951): T141–T144.

Hodge, J.E., “Chemistry of browning reactions in model systems,” J. Ag. and Food Chem.1 (1953): 928–940.

Hutchins, J.K., The water-soluble components of degraded cellulose (MS thesis), North Carolina State University1981.

Kerr, N., The degradation of cellulose: causes and prevention(Final Report to the Administrator of the National Museum Act on Grant No. FC-9059580000 [79/245]), 1980.

Madaras, G.W., and Turner, H.A., “Further observations on the effects of evaporating water from cotton cellulose,” J. Soc. Dyers and Colourists69 (1953): 371–377.

Meynell, G.G., and Newsam, R.J., “Foxing, a fungal infection of paper,” Nature274 (1978): 466–468

Mithel, B.B., Webster, G.H., and Rapson, W.H., “The action of water on cellulose between 100 and 225C,” TAPPI40 (1957): 1–4.

Press, R., “Observations on the foxing of paper,” Int. Biodetn. Bull.12 (1976): 27–30.

Rollinson, S.M., “The colored water-soluble materials of heated bleached kraft pulps,” TAPPI38 (1955): 186–192.

Schaffer, K., Appel, W.D., and Forziati, F.H., “Reactions at wet-dry interfaces on fibrous materials” (Research paper 2750). J. Res. Nat. Bur. Standards54 (1955): 103–106.

Strachan, J., “Solubility of cellulose in water,” Nature141 (1938): 332.

Section Index

Copyright 1983 American Institute of Historic and Artistic Works