All teeth were preliminarily cleaned in an ultrasonic bath for fifteen minutes, and then dried for five hours at 40° C.
The enamel of teeth coming from an archaeological context is extremely fragile, mainly due to the dehydration of dentine resulting in a loss of its characteristic resiliency. Consequently, the specimens were embedded in plastic material before being sectioned. The expoxy resin (Buehler Ltd.) was used with excellent results. The teeth were placed in an ice cube tray treated with a nonstick release agent (Buehler Ltd.). It was not necessary to pay close attention to the positioning of the tooth in the mould space, as the holder of the microtome permitted a great deal of flexibility in the orientation of the blocks to be sectioned. After the resin was poured into the mould containing the teeth, it was immediately placed in a vacuum desiccator, pumped down, and left until all bubbling of the resin ceased (around 10-15 min).
The resin was left to cure for 24 hours at room temperature, and then the blocks were extracted from the mould. With respect to protocol provided by Marks and co-workers (1996), we found that the Epo Kwick resin was stable and adhered well to the enamel. Before starting the cutting sequence, it is preferable to age the blocks for one week.
One, or even two, longitudinal bucco-lingual slices (400 m m thick) were cut from the midsection of each tooth using a diamond blade microtome (Leica 1600, Leica AG), and were affixed to slides with a mounting medium (Eukitt, O. Kindler GmbH & Co.).
With respect to the incisors and canines, special care was taken to cut the sections as close to the central axis of the tooth as possible, and the molars were sectioned through at least one buccal cusp. If two sections were available, measurements were generally taken on the section with the smallest occlusal thickness of the enamel (thus, the most central section). It was not always possible to obtain perfectly centred sections, so in these cases any morphometric parameters in which the measurements would have been biased were not recorded.
Thin sections were ground on a motorised grinder (Minimet 1000, Buehler Ltd.) using two abrasive pads of different grits (Carbimet 320 and 1200, Buehler Ltd.). Final polishing was done with a microcloth (Buehler Ltd.) and alumina (0.05 µm Micropolish B alumina, Buehler Ltd.). The final thickness of the sections ranged between 70 and 150 µm. After dehydration by means of three rinses in alcohol at different concentrations (70%, 90%, 100%), the thin sections were mounted on glass slides with Eukitt.
