Accurate determinations of abundances have been hampered by the so called t^2-problem. This is probably one of the most important problems in present day PN research. The t^2-parameter is a measure for the degree of electron temperature (Te) fluctuations in the ionized region. Such fluctuations increase the strength of emission lines. This effect is stronger for lines originating from higher energy levels. Consequently, if optical line ratios (e.g. [O III] 4363,4959,5007) are used to calculate Te, and if the fluctuations in Te are present, the result will be systematically too high. Since Te is used to convert ionic abundances to elemental abundances, the t^2-parameter has a substantial influence on abundance determinations. Models do not predict fluctuations in Te, but observational evidence gathered so far contradicts this. The first goal of this proposal is to increase the small amount of observational data currently available and thus to obtain a stronger basis for investigating the t^2-problem. The second goal is to determine accurate abundances for the PN in our sample. The abundances in Planetary Nebulae (PN) reflect the evolutionary history of the central star. Especially Carbon and Nitrogen are affected by its chemical evolution. Evolutionary models predict correlations between the core mass of the star and the ratios He/H, C/O and N/O. However no clear observational evidence for such relations exists and it is unclear whether this is caused by errors in the models or the abundance determinations. We want to use the ISO data to search for possible correlations between the above mentioned abundance ratios and luminosity, which is directly related to the core mass. In order to do the analysis we need accurate measurements of mid- and far- infrared fine-structure lines, and the IR continuum shape. These observations cannot be done from the ground. Only ISO provides the required signal-to-noise to reach our goals.