Biostratigraphical investigations of Lago d'Averno, near Naples

Remains of diatoms, molluscs, ostracods, foraminifera and pollen exines preserved in the sediments of Lago d'Averno, a volcanic lake in the Phlegrean Fields west of Naples, allowed us to reconstruct the changes in the ecological conditions of the lake and of the vegetation around it for the period from 800 BC to 800 AD. Lago d'Averno was at first a freshwater lake, temporarily influenced by volcanic springs. Salinity increased slowly during Greek times as a result of subsidence of the surrounding land. Saline conditions developed only after the lake was connected with the sea by a canal, when Portus Julius was built in 37 BC. The first post-Roman period of uplift ended with a short freshwater phase during the 7th century after Christ. Deciduous oakwoods around the lake was transformed into a forest of evergreen oaks in Greek times and thrived there - apparently almost uninfluenced by man - until it was felled, when the Avernus was incorporated into the new Roman harbour in 37 BC, to construct a shipyard and other military buildings there. Land-use was never more intense than during Roman times and weakest in Greek and Early Roman times, when the Avernus was considered a holy place, the entrance to the underworld.

The effect of boundary conditions in the numerical solution of 3-D thermoelastic problems

After the extensive research on the capabilities of the Boundary Integral Equation Method produced during the past years the versatility of its applications has been well founded. Maybe the years to come will see the in-depth analysis of several conflictive points, for example, adaptive integration, solution of the system of equations, etc. This line is clear in academic research. In this paper we comment on the incidence of the manner of imposing the boundary conditions in 3-D coupled problems. Here the effects are particularly magnified: in the first place by the simple model used (constant elements) and secondly by the process of solution, i.e. first a potential problem is solved and then the results are used as data for an elasticity problem. The errors add to both processes and small disturbances, unimportant in separated problems, can produce serious errors in the final results. The specific problem we have chosen is especially interesting. Although more general cases (i.e. transient)can be treated, here the domain integrals can be converted into boundary ones and the influence of the manner in which boundary conditions are applied will reflect the whole importance of the problem.