The knowledge of the chemical composition of metallurgical components has become of increasing and sometimes v ital importance. The behaviour and effects of. for example, the heat affected zone adjacent to welds, joins in dissimilar metals. corrosion products. overheating and embriulement all depend critically on chemical composition (e.g. metallurgical phase changes during embrittlement). With existing, older equipment, chemistry is often unknown. In nuclear industries the plant is likely to be radioactive and cannot be accessed. Current methods of industrial analysis are testricted to cutting out material from the metal and submitting it to more traditional chemical assay. However. this approach has severe limitations on highly stressed high temperature components. Tt is the proposal cf this project that it is possible to make laser spectroscopy practicable and viable for use as a remote industrial diagnostic tool by specifically addressing its limitations and applying new and innovative technologies to overcome them. The project intends to develop the following in order to achieve this goal: * Advanced sig noise and data analysis systems to obviate the current limitation on the temperature of the component being analysed (up to component te the range 500<°C<1OOO). * Automatic electronics for control of positio (where current systems use manual methods which greatly increases its complexity of use whilst limiting its reproducibility). * Improved so (much current software. in conjunction with the spectroscope. limits the resolution to which the spectrograph can separate emission lines. Cu software is also particularly user-unfriendly). * Selective spectrometers. * Tuneable lasers for particular spectral line excitation. Until now such an approach has not been taken to laser spectroscopy and. as a result, has totally precluded its potential as a v ery powerful and beneficial industrial diagnostic tool.