Very few bones survived 5 weeks exposure. Sixteen bones were recovered from the upper soil level by hand and wet-sieving; none from the 50 cm square quadrat (see Table 1).
A total of 13 herring bones were recovered. This represents less than 1% of the original bones. Only three mackerel bones were found; roughly 1% of the bones laid out. These calculations are based on there being 80 'major' elements per herring and 50 per mackerel (Jones 1986).
It is difficult to predict which of the 'major' elements survive best. In this experiment it was the maxilla, supramaxilla, metapterygoid, and caudal vertebrae of the herring and the premaxilla, dentary, and caudal vertebrae of the mackerel which survived.
While it is not possible to be certain of all the mechanisms contributing to the loss of bones from the experimental area, there is some evidence of the fate of the bones. A large well-fed labrador dog living close to the site was seen to eat some of the fish waste. Local inhabitants of the valley suspected that mink Mustella vison, and rats, Rattus norvegicus may also have removed some fish.
The condition of the recovered bones is of interest. Most of the identifiable elements were complete, showing that little damage had occurred in situ. No unidentified fragments were recovered. We can conclude that scavenger damage to bones occurred off the site and that large numbers of whole bones (probably articulated to other whole bones) were removed from the site.
Both head elements and vertebrae were recovered, suggesting that whole fish were present. It is interesting to note, however, that no herring cleithra were recovered (40 were laid out). Seeman (1986) has convincingly argued that the absence of herring cleithra at a Dutch whaling station on Spitzbergen indicated that the herring from that site were cured by the 'Dutch' method. This involved the removal of the cleithrum branchiostegals and associated elements with internal organs prior to preservation. Such a suggestion for the material from Castell Henllys would be erroneous.
Considering the recovered assemblage as an archaeological sample, and using it to reconstruct the original quantity of fish present at the site, illustrates the difficulties encountered when interpreting archaeological deposits. The minimum number of individuals present, as calculated, was one herring and one mackerel; there were in fact twenty herring and five mackerel.
It is interesting to consider the number of bones recovered from each species and to relate these to the original numbers of identifiable elements laid out. Of the bones recovered, 81% are herring and 19% are mackerel; the original assemblage comprised 86% herring and 14% mackerel. In this experiment, the relative proportions of observed and expected elements appear to have been little affected by the factors contributing to taphonomic loss.
© Internet Archaeology
Internet Archaeology
Last updated: Tue Jul 7 1998