Achaeometrical Research in Hungary II., 1988

ANALYSES - POTTERY - György SZAKMÁNY: Insight into the manufacturing technology and the workshops: evidence from petrographie study of ancient ceramics

of the clasts and estimate the number of frequency maximum they have. Reporting the average grain size of each individual grain size group is useful in the interpretations of tempering and manufacturing technology. A similar role can be played by maximum grain size as well. Average grain sizes in Roman amphoras may exceed 150-200um in diameter (JÓZSA and SZAKMÁNY, 1987) but there are some types of amphoras with very small grain sizes average (<50um). The average grain size of Neolithic potsherds is coarser than those of the ceramics mentioned above (SZAKMÁNY, 1996). Some individual carbonate grains, for example, reach or exceed 1mm in diameter in such Neolithic potsherds. Furthermore, in the Roman amphoras (except for some Hispanic Roman amphoras) the maximum grain size occasionally exceeds 500цт in diameter. The distribution of grain size also is highly variable. Frequency histograms may show 1, 2, or 3 or even more peaks. However, the area(s) between the individual maximum values should be carefully studied, in order to find out whether the distribution of grains are well-sorted or unsorted (Figure 1). The lack of grains between two peaks may indicate that the tempering material was mixed with clay twice, or the tempering material had originally possessed this character which is typical of fluviatile sands. If the clasts are completely unsorted, it may mean that the tempering matter had been crushed before it was mixed with clay. Usually, in this case, the shapes of grains are mostly angular. It is also strongly recommended to give an approximate quantitative estimate of individual grain sizes. The degree of roundness of clasts is generally not too useful in studying ancient ceramics. If the grains are perfectly angular or angular (according to the roundness scale of POWERS, 1953), then crushed rocks or coarse grained sands were mixed in the clay as a tempering material. If the temper derived from fluviatile or shoreline sands, the roundness of grains depended on the roundness of the clasts in the original sands. These latter source materials are generally subrounded and rounded. Soft clasts (e.g. carbonates) are more rounded than the others. Therefore, the observation of any difference in the degree of the roundness of clasts is significant for the purposes of interpretation. The porosities of ceramics are variable. Primary and secondary porosity can be distinguished. It is believed that primary porosity formed during the drying of ceramics. Secondary porosity, in turn, developed during the firing procedure. If the raw materials contained iron-oxide (e.g. magnetite, hematite) and the firing took place at high temperatures and in an oxidysing atmosphere, the iron-oxide underwent dissociation. This process yields free oxygens which can lead to the secondary porosity. Under reducing conditions and low temperature, on the other hand, iron-oxides are reduced due to the presence of carbon-monoxide and carbon-dioxide before the melting mechanism. Therefore, under the subsequent high T conditions, free oxygens are absent and secondary porosity cannot develop (KAKASY and SOMODI, 1979). In summary, the porosity illustrates mainly the firing conditions. It should be noted, however, that firing conditions depend in part on the kiln in which the ceramics were placed during the firing. For example, the majority of the Roman amphoras (namely type Dressel 6B) are massive, containing only a few pores, whereas the subgroup of "Laek Bass" from this type of amphoras has a considerably higher porosity (JÓZSA and SZAKMÁNY, 1987). Conclusions 1) The study of the ancient ceramics using pétrographie microscope provides considerable information concerning the origins of ceramics and their manufacturing technologies. 81