JAIC 1993, Volume 32, Number 3, Article 1 (pp. 213 to 230)
JAIC online
Journal of the American Institute for Conservation
JAIC 1993, Volume 32, Number 3, Article 1 (pp. 213 to 230)





Several samples from the Walters objects were examined and compared to the known sample of baleen discussed in section 2. Samples of leather and horn were also prepared for comparison. All samples were mounted in Cargille Meltmount (n = 1.662) and viewed with transmitted light under crossed and uncrossed polars using a Leitz analytical microscope. The samples prepared included: (1) a lateral section of the sword hilt wrappings; (2) a lateral section from the helmet battle axe handle; (3) a known sample of worked cattle horn; and (4) a known sample of calf leather.

Samples 1 and 2 were examined. Both samples looked similar to the known baleen sample and showed little structure at 250x. In transmitted light, the sections were pale tan colored and translucent; dark spots were visible in areas, and there was some indication of direction. When viewed under crossed polars both samples exhibited a full range of polarization colors; thinner areas were restricted to white, black, and gray. In both cases a faint “herringbone” or braided pattern was visible at points of extinction. In sample 1, a slightly thicker sample, a faint linear pattern was visible under crossed polars parallel to the length of the sample. Under the same conditions the known baleen sample exhibited a vivid rainbow of polarization colors and a faint “herringbone” pattern at extinction. The horn sample (sample 3), also showed little structure and under crossed polars exhibited a range of blue, pink, black, and white colors. Under crossed polars at points of extinction, banding parallel to the length of the sample was visible; the banding in this case was far more clearly delineated than that seen in sample 1. A “herringbone” pattern was also visible every 90 at points of extinction. The leather sample (sample 4) contained large fiber bundles and open pores; it exhibited orange, white, and tan polarization colors and looked completely different from both the baleen and the horn.

It appears that if the Walters sword hilt wrappings are baleen, they were not made from a complete cross section but rather from the horny outer covering of the baleen plate; thus, the positive identification of these sections is very difficult. To help confirm the identification, a sample from one of the sword hilt wrappings was sent to David Stoney of the University of Illinois, who is conducting research on the identification of baleen versus rhinoceros horn. His work thus far is based on optical microscopy but may later include scanning electron microscopy. Given the lack of structure in the sample, Stoney was unable to confirm or rule out the identification of baleen. It was evident from the lack of tubular structure that the wrappings definitely were not made from the fibrils or hairlike structure of a baleen plate. During his examination Stoney observed only muted black, white, and blue polarization colors.

According to Barbara Stephan, who had been working in conjunction with Stoney and is now affiliated with the McCrone Laboratories, Chicago, polarization colors are always present in baleen, but the particular colors depend on the species of whale, the thickness of the sample, and the pigmentation of the baleen itself. In very thin samples only first-order polarization colors—black, gray, and white—are visible. As the sample thickens, second-, third-, and fourth-order colors are visible to include a full range of the spectrum. According to Stephan, finback baleen exhibits the most vivid range of polarization colors, while bowhead baleen, which can be heavily pigmented, shows only muted colors.

In order to see whether boiling and shaping baleen affects its optical properties, two samples of finback whale baleen were cut from the outer and inner edges of a modern plate, two areas in which the texture and thickness of the baleen are quite different. Cattle horn was also sectioned for comparison. The tests conducted in the lab were not meant to mimic traditional techniques but were only quick exercises to get a feel for how long it would take to soften baleen.

Each sample was placed in a beaker of distilled water and boiled for 45 minutes. The samples were then removed from their water baths and tested for flexibility. Sample 1 (thin outer portion of plate) and sample 3 (horn) were very flexible and felt like a soft fingernail. Sample 2 (thick inner portion of plate) was still somewhat stiff, and some of the individual fibrils had begun to delaminate. The samples were then allowed to soak overnight in room-temperature water. In the morning the samples were removed again; samples 2 and 3 had become quite stiff, but sample 1 remained quite flexible. The samples were then boiled again for 45 minutes and pressed between bent metal mending plates and allowed to dry. The samples were removed after 3 days; each retained the shape of the mending plates and took on a slick plastic look from the surface of the metal.

Longitudinal sections from the interior curvature of each molded sample were mounted and viewed under polarized light. The appearance of these samples differed very little from the samples taken before “boiling and shaping.” This finding implies that working in this manner has little effect on the optical properties of the baleen when it is viewed under polarized light.


Other basic tests were performed in the laboratory to help identify the baleen. Several known samples of baleen were viewed with long-wave ultraviolet light side by side with the unknown sample from the sword hilt and a sample of horn. The baleen samples fluoresced to differing degrees depending on their degree of pigmentation. The horn sample was very translucent, sparsely pigmented, and fluoresced bright yellowish-white. The unknown sample from the sword hilt was dark black-brown in visible light and did not fluoresce. The boiled and molded samples of finback baleen were then viewed with UV light and compared with the original baleen plate from which they were taken. The baleen plate itself fluoresced pale yellowish-white. The thick sample taken from the inner margin of the plate looked the same in UV light after treatment, but the thin sample taken from the outer margin of the plate no longer fluoresced in the same way but now appeared dull brownish-orange. When the water baths were reexamined after soaking, some residue remained in the bottom of each beaker. The loss of this solubilized material may have been responsible for the change in appearance of the thin baleen sample under ultraviolet light. A sample from each beaker was saved for future analysis.

When heated with a flame, baleen smells like burning hair. Other keratin-based materials such as wool and horn also exhibit this characteristic, so it alone cannot be used as a definitive test for baleen. Both unknown samples from the sword hilts and the helmet battle axes and a sample of finback baleen were heated with a bunsen burner, and each smelled like burning hair.

Toward the end of the project, Mary Baker of the Smithsonian Institution Conservation Analytical Laboratory carried out Fourier-transform infrared analysis on two samples: one from a plate of baleen and one from the Japanese sword hilt wrappings. The baleen produced a clear spectrum for keratin. The unknown sample produced a very similar spectrum with some peak shifts, probably due to subsequent working of the baleen.

In light of this investigation, there appears to be little in the way of simple laboratory tests to clearly distinguish horn from baleen. The identification of baleen as the material used for the sword hilts and the substrate of the battle axe handles on the Japanese helmet was based upon the similarities of their physical and optical properties to those of known baleen samples and the knowledge that baleen was used in Japanese arms and armor. Because horn is much shorter than a baleen plate, the size of the object or part thereof may also help in identifying the material. The sword hilt wrappings consist of strips several feet long, which rules out the possibility of horn. The battle axe handles from the Japanese helmet are also over a foot long, and each is approximately 6 in wide. Their wide, flat shape is reminiscent of a baleen plate.

Because baleen and horn are both hard keratins and exhibit so many similar characteristics, continued research is necessary to help distinguish the two. In this study only cattle horn was compared. Until further research is carried out, form, function, and the availability of the material to a particular culture may be most helpful in distinguishing between horn and baleen.

Copyright 1993 American Institute for Conservation of Historic and Artistic Works