Dr. Corina Rogge, the Andrew W. Mellon Research Scientist at the Museum of Fine Arts, Houston, presented a very interesting (and thorough) talk on the “Analysis of Historical Tintype Plates: Materials, Methods and Manufacturers” giving a brief overview of the tintype process and the ongoing project to characterize their material properties. The analysis was undertaken on a study collection involving 226 unknown tintypes.
Despite the popularity of the process, there were only three patents filed in its history, and how they were actually made is poorly known. Hamilton Lamphere Smith and Peter Neff first patented their process for a collodion image on a japanned surface in 1856. Their melainotypes, so named from the Greek melano, or black, consisted of a varnished image layer on the japanned iron plate, also varnished verso. Peter Neff was interested in the commercial properties, including japanning purchased plates. He sold prepared plates, as well as licensing the process for use. The plate itself was not patented. Victor M. Griswold patented a competitive process also in 1856, using bitumen on an iron plate, and called them ferrotypes to indicate the iron content. There were eventually several plate manufacturers including Neff, Griswold, Dean & Co., and Phenix; however, only Neff and Griswold stamped their plates.
Tintypes were affordable, durable and accessible, and became widely popular in mid-1800s America, and really came to represent the middle class. They are not as common in museum collections, but a large number survive as family and historical items. Dr. Rogge noted that the market and development of the process were strongly economically driven, and the producers found a need to create trends to keep their edge or drive sales. For example, Griswold began to name and sell imperfect plates as “eggshell,” in addition to glossy plates.
Dr. Rogge’s study sought to determine whether different manufacturers used different materials, and if so, whether they be differentiated and dated. X-ray fluorescence spectroscopy (XRF) analysis revealed mostly iron, as would be expected for the plate. A smaller number contained manganese, but didn’t seem to correspond to a trend. There were no major differences in thickness, aside from very early Neff plates being thicker. These very were fairly quickly thinned down to the typical size, making them cheaper to manufacture and mail. Griswold used thicker plates only very briefly.
Organic analysis was carried out on the japanning layer by pyrolysis gas-chromatography (py-GC-MS). There were no particularly unusual result, as might be expected for a commercial process. Drying oil was the most common component, then pine resin, followed by a small number containing shellac. Asphaltum, which Dr. Rogge pointed out is difficult to work with, was found in fewer than 10% of the plates although it is mentioned in two of the three patents ever filed.
Microscopy determined the most common pigments of the japanning layer to be carbonaceous blacks, as given in the Neff patent, and iron earth pigments as given in the later patent. Plates were sometimes advertised as chocolate or black, and an effort was made to differentiate between the two. Colorimetric analysis ignored the CIE b* parameter (blue-yellow) to accommodate the assumed yellowing of the varnish. CIE a* when checked against manganese didn’t correlate. In attempting to differentiate the bare support versus japanning layer with support, the thickness of the japanning layer seemed to have much more influence. However, Dr. Rogge believes further study may prove chocolate plates (more red) correlate to the use of iron earth pigments in the japanning layer.
Relatively few plates had a single japanning layer, while many had two or three. Though the practice is not indicated in any of the patents, it is mentioned in articles of the time on how to make tintypes. Though multiple coats would increase the cost and difficulty of production, they would provide a smoother, glossier surface. In addition, a more pigmented bottom layer (a ‘face coat’) might serve to isolate and protect pigments from reactive image silver, and approximate the depth of an ambrotype by allowing more light to pass through.
Dr. Rogge pointed out that the project needs more data but points to the possibilities of trends and potentials for identification. She noted also that the actual plates varied significantly over time, and differed between manufacturers and particularly from the patented processes. For example, Neff plates may have more layers as time goes on, and Griswold plates may be associated with profiles containing asphaltum, drying oil, pine oil and shellac. Such an ability to profile the plates would give the possibility of identifying and dating unknown plates.