The aim of this project was to locate and characterise sites of plasticity involved in long-term memory in a model invertebrate system, the snail Lymnaea stagnalis. Plastic changes are likely to involve integrating neurons in the sensory pathways that process the conditioned and reward stimuli used in chemical appetitive conditioning. The cerebral ganglia of the Lymnaea brain are a likely location for sensory integration making them a primary target for this investigation. A chemosensory nerve innervating the cerebral ganglia was dye-filled together with a nerve linking the cerebral ganglia to the feeding network. These experiments revealed specific sites of potential synaptic contact between the two nerves and subsequently, six new cerebral-buccal interneurons were identified and characterised electrophysiologically. Defining nitric oxide synthase distribution in peripheral and central neurons provided another route to finding sites of plasticity since nitric oxide is required for longterm memory formation in Lymnaea. Electrophysiological correlates of behavioural learning were found in the feeding motoneurons and the connective containing the cerebral-buccal interneurons. Variations in conditioned stimulus concentration and sites of its perfusion were discovered to be of crucial importance for the observation of these learning correlates. Determining the role of the CA1, CT2 and CV1 a interneurons in conditioned responses was a central aim. Evidence for their possible role in learning included: their response to the conditioning and reward stimuli in naive, conditioned and control animals; their role as modulatory interneurons of the feeding network and their anatomical and electrophysiological connectivity to primary sensory neurons and interneurons of the feeding network. A change in the response of the CA1 cell to the reward stimulus after conditioning was thought to arise due to pre-exposure to the conditioned stimulus. No individual neuronal change was found that could account for conditioned feeding responses in the whole network