The effects of adenosine agonists and antagonists were examined on population excitatory postsynaptic potential (population EPSP) slopes, population spike (PS) amplitudes and the relationship between the two i.e. EPSP-spike (E-S) coupling. Activation or blockade of adenosine A2A receptors responses, evoked by stimulation in the stratum radiatum, had no effect on these parameters. However, activation of the adenosine A1 receptor using N6-cyclopentyladenosine (CPA) resulted in a significant decrease in both population EPSP and population spike. When adenosine A1 receptors were activated using CPA (50 nM) the decrease in population spike amplitude was greater than could be accounted for by the decrease in population EPSP, resulting in a dissociation in the EPSP-spike relationship as measured by a right-shift in the E-S curve or a decrease in the ratio PS/EPSP. When adenosine A1 and A2A receptors were activated at the same time using CPA and 2-[p-(carboxyethyl)-phenylethylamino]-5'-N-ethylcarboxamidoadenosine (CGS 21680) respectively, the depression of the E-S relationship was significantly attenuated. This was due to a mainly postsynaptic effect whereby activation of A2A receptors significantly attenuated the postsynaptic elevation of action potential threshold by adenosine A1 receptor activation. The use of a variety of signal transduction modulators suggested that adenylate cyclase, protein kinases or nitric oxide were not involved in the interaction between A2 and A1 receptors. Blocking potassium channels with barium or glibenclamide reduced the inhibitory effects of CPA on population spike amplitude but not population EPSP slope and attenuated the inhibitory effects of adenosine A1 receptor activation on E-S coupling in a similar manner to adenosine A2A receptor activation. The increases in the ratio PS/EPSP seen when A2A receptors were activated with CGS 21680 or when potassium channels were blocked with barium were not significantly different, suggesting that a change in postsynaptic potassium conductance may underlie the antagonistic effect of adenosine A2A receptor activation on A1-mediated responses. Application of ATP 10 μM to hippocampal slices initially decreased population spike amplitude and then produced long-term potentiation (LTP). 2.5 μM ATP also resulted in LTP of responses although did not inhibit the potentials during perfusion. The stable analogue of ATP, αβ-methyleneATP had no long-term effects on potentials. Blocking P2 receptors (with suramin) or N-methyl-D-aspartate (NMDA) receptors prevented the induction of LTP. When adenosine A1 receptors were blocked or the slices superfused with adenosine deaminase no inhibition of responses was seen during perfusion with ATP, nor was LTP induced. Adenosine itself did not induce LTP. These results suggest that induction of LTP following ATP perfusion requires the activation of both P2 and A1 receptors. During superfusion with ZM 241385, ATP 10 μM still induced LTP but no longer caused the inhibition of responses during perfusion. Following electrically induced LTP (100 Hz for 1 second), perfusion of ATP caused no further potentiation of responses, suggesting that electrically and ATP induced LTP share common mechanisms. Perfusion of a preparation of 5'adenylic acid deaminase (AMPase) from Aspergillus species for 10 min resulted in an initial increase in population spike amplitude which gave way to a decrease in potential size and EPSP slope that remained depressed for at least 30 min following washout of the enzyme, resulting in long-term depression (LTD). A decrease in EPSP-spike coupling was observed 30 min following washout of enzyme. Superfusion of AMPase on slices in which transmission had been potentiated using high frequency stimulation resulted in depotentiation of responses. The LTD caused by AMPase could not be inhibited by the allosteric inhibitor of the enzyme 2,3-diphosphoglyceric acid. Further investigation also ruled out the involvement of nitric oxide, protein kinase, and cyclooxygenase in the activity of the AMPase. The use of AMPase extracted from rabbit muscle resulted in no LTD of responses. Analysis of the crude enzyme extract from Aspergillus revealed that the preparation was not pure. Separation of the various components in the crude preparation led to the discovery that the AMPase was not the active factor causing depression of evoked responses. Further purification of the active protein and subsequent sequencing showed that the protein was similar to a β-glucosidase. However, perfusion of β-glucosidase extracted from almonds did not result in LTD of responses. Overall, this thesis extends knowledge of how synaptic transmission can be modulated by purines, and raises the possibility that a new family of proteins, related structurally to glucosidases, may also affect synaptic plasticity.