The processing of hammerscale samples taken during excavations at Cardiff Castlehas revealed some interesting and surprising finds (further details on the excavation can be found within GGAT Newsletter 1 in our newsletter archive.
The hammerscale samples were taken from an area next to a Roman road in which evidence of industrial activity, such as kilns and pits of hammerscale and slag were found. Hammerscale are the scale-like fragments of metal, which break off when striking a heated metal object with a hammer in ordered to shape it.
In order for the hammerscale samples to be analysed by a specialist they must first undergo wet sieving. Initially the samples are submerged in a bucket of water and swirled around by hand to loosen the hammerscale from any other material in the sample. Then the mixture is poured through two interlocking sieves placed one above the other, the sieve with the finest mesh being at the bottom to collect the smaller scales, leaving the lager scales and other objects of similar size (including stones) in the top sieve. The process is helped by rubbing the finer material through the upper sieve with fingertips, much in the same way as you would when sieving flour but a lot rougher on the hands and nails! Pouring water into the upper sieve helps the finer scales to pass through the mesh to be collected on the lower sieve. Once this stage is complete the material in each sieve must be kept separate and dried in order to be sent to a specialist for further analysis.
During the wet sieving of these samples some burnt bone fragments, a copper bead and Roman pottery were found. The pottery sherds include Samian, Black-burninish, and greyware, one piece of which is decorated with an incised design. But perhaps the most intriguing find was a bright green tooth that appears to be human! It is open to debate how the tooth came to be there, but it is likely that the unusual colouring was caused by the copper waste. The tooth and pottery sherds are particularly interesting as they came from an area known to be associated with the earliest Roman fort at the site, dating to the 1st century. We are looking forward to the results from the specialist’s analysis, so watch this space!