4 DISCUSSION

Of the methods used, measuring with the glossmeter proved to be the most quantifiable method for recording changes in the surface of the samples. These changes included the development of cracks, but also roughening, minute dimpling, and other small deformations corresponding to the grain of the paper. Reduction in gloss most likely results from the stress imposed on the albumen layer during the wetting and drying cycles. The strain in the albumen layer is relieved by cracking and/or surface roughening, the degree of which depends on two factors: (1) the amount of stress imposed by the specific treatment, and (2) the differing responses of the albumen layer and the paper substrate.

Samples bathed in the water:ethanol mixture exhibited the least reduction in gloss. They were exposed to the smallest amount of water and, as a result, probably experienced the smallest degree of wet-to-dry expansion and contraction. Conversely, samples bathed in water followed by ethanol showed the greatest amount of gloss reduction and roughening of the surface. In this treatment, the samples expanded during the water bath and remained swollen as the water was displaced by ethanol. However, ethanol evaporates rapidly, causing dimensional changes that may have resulted in the observed increased gloss reduction. Furthermore, the displacement of water by ethanol may desiccate the emulsion, encouraging brittleness. The samples bathed in water and alkaline water showed an intermediate reduction in gloss. These samples may have expanded to a degree similar to those in the bath of water followed by ethanol but were not subjected to the rapid drying caused by immersion in ethanol.

The thicker samples consistently exhibited a greater reduction in gloss than those with thinner emulsions. Examination with the stereo-microscope at 40x provided general information about the number, size, and orientation of the cracks formed during this treatment study. In general, a greater number of cracks formed in the thicker samples than the thinner ones. Additionally, the cracks in the thicker samples were relatively long (>0.5 mm) and tended to form across the grain, while those in the thinner samples tended to be shorter (<0.1 mm) and showed no preferential direction. Similar observations have been reported by Alice Swan (1981, 275):

The severity of the fissuring and cleavage appears most directly related to the thickness of the albumen layer. The only albumen prints I have seen completely without fissuring are prints from the 1850s and 1860s, so thinly coated that they have almost no gloss and are sometimes mistaken for salt prints. Yet a more thickly coated spot on such a print will show the usual fissuring and cleavage, and the amount and severity will depend on the thickness of the spot.

These observations suggest that more stress may be assumed during treatment cycling by thicker layers of albumen than by thinner ones, which are better able to follow the expansion and contraction of the paper substrate.