6 CONCLUSIONS

6.1 GENERAL

Excursions to high relative humidity and wetting swell the albumen more than it can stretch when it is dried. When it shrinks during desorption, its inherent low tensile strength results in stress relief cracking of the albumen.

A layered structure in albumen prints and a mismatch in dimensional response of the layers to water vapor result in albumen print curl and planar distortions. The albumen/paper composite is stronger than the print, and it swells and shrinks more than the paper alone.

At high rates of dimensional change, albumen behaves as a stiff brittle solid. At a slow rate of strain, on the order of 1.5% stretch in 12 minutes, it behaves as a rubber. Thus, surface swabbing with water and large climatic changes have the same ultimate result, even though the step-by-step actions are different.

6.2 STORAGE CONDITIONS

The mechanical and dimensional behavior of albumen photographs underscores the importance of a controlled storage environment with constant, moderate RH. A 30% or greater change in RH centered on 50% RH (between 35% and 65% RH) will exceed the strain limits of an unsupported albumen film at slow rates of change and will result in crack width increase or the introduction of new cracks. Parenthetically, 50% RH has no storage-specific importance; rather, it was the environment used for testing.

6.3 TREATMENT

The materials properties findings have implications for the conservation treatment of albumen photographs.

Water vapor treatments exceeding 30% (50 � 15% RH) and the direct application of liquid water result in swelling and shrinkage that will cause stress-release cracking.

The most damaging stage of the wetting and drying process is the shrinkage during drying. During wetting the materials are plasticized and can yield to the forces of compression. In the case of the historical albumen print in the E-SEM experiment (Messier and Vitale 1993), the albumen layer should have swelled 17%+ in all directions, but it was restrained in the x and y planes. It swelled 5% and then shrank roughly 9%. Thus additional damage occurred during drying. In addition, the curling in the albumen layer occurred during the drying cycle.

Rate of drying can influence crack population increase and crack propagation behavior. Drying an albumen print, after it has been immersed, usually occurs over 1 or more hours. For a print that swells 5%, an increase in crack population and crack propagation will occur because, no matter what the rate of shrinkage, cracking will result because the swelling is greater than 1.5%. If the same print is dried from immersion in less than 48 minutes, more cracks and greater crack width increase than those reported in Messier and Vitale (1994) should be observed because of the pronounced rate-of-strain dependency of the material. Slowing the drying time, by prewetting the blotter for example, may have no effects on crack width or crack population, but certainly there would be no detrimental effects due to a faster rate of shrinkage in this rate-of-strain–dependent material.

6.4 NEED FOR UNMOUNTING OR MOUNTING

The long-term mechanical and physical effects of mounting albumen prints onto rigid supports need investigation. Is a mounted albumen photograph better preserved (in terms of the stability of the albumen layer) if the print is left on its mount? Would dry removal of a mount be preferable, or would the mechanical action also result in additional cracks in the albumen? The rate of strain during mechanical removal would, in most cases, surely be in the fast domain. Would albumen photographs that have never been mounted be more stable if they were adhered (using nonaqueous adhesive) to a rigid mount? If an albumen print is removed from its mount during the course of a treatment, is it better to return it to a rigid support, or should it remain unmounted?

ACKNOWLEDGEMENTS

The authors wish to thank the following for their assistance in preparing this work. Duane Chartier, private conservation treatment and conservation science practice in Los Angeles, offered valuable assistance in uncovering the importance of rate-of-strain on albumen—an excellent resource. Donald Hunston, Polymers Division, National Institute of Standards and Technology, provided the use of ultrasonic impediometry equipment and interpretation of the data.

Marion Mecklenburg, assistant director for conservation research at the Conservation Analytical Laboratory (CAL), Smithsonian Institution, provided insightful guidance, read the manuscript, and made many valuable contributions. Mark McCormick-Goodhart, photographic scientist, CAL, offered valuable advice and insight at many points in the project. Noreen Tuross, research biochemist, CAL, performed amino acid assay on two samples of albumen, before and after denaturing. Carol Grissom, chief of objects conservation, CAL, proofread the manuscript and made many valuable suggestions. Gil Taylor, chief librarian at the Museum Support Center Library, a division of the Smithsonian Institution Libraries, deserves our warmest appreciation for providing valuable bibliographic information and searches and for tirelessly assisting in the acquisition of articles and books on this project.