A simple hypothesis can explain the results obtained to date if we disregard those results when we wait 30 minutes after original learning to inject. The hypothesis is that, as a result of learning, the postsynaptic endings at a specific set of synapses become more sensitive to transmitter. This sensitivity increases with time after initial learning and then declines. The rate at which such sensitivity increases depends on the amount of initial learning. If the curve of transmission plotted against time is displaced upward with anticholinesterases then the very low portions will show facilitation, and the high portions will cause block (Fig. 8). The middle portions will appear unaffected (unless special experimental tests are made). If the curve of transmission is displaced down with anticholinergics, then the middle portion will appear unaffected and only the very early or late components will show block.
The results are evidence that synaptic conductance is altered as a result of learning. So far it seems (i) that cholinergic synapses are modified as a result of learning and that it probably is the postsynaptic membrane that becomes increasingly more sensitive to acetylcholine with time after learning, up to a certain point. (ii) After this point, sensitivity declines, leading to the phenomenon of forgetting. (iii) There is also good evidence that there is an initial phase of declining sensitivity to cholinesterase or increasing sensitivity to anticholinergics. This could reflect the existence of a parallel set of synapses with fast decay that serve as a shortterm store. (iv) Increasing the amount of learning leads to an increase in conductance in each of a set of synapses without an increase in their number. (v) Both original learning and extinction are subserved by cholinergic synapses.