Search

The projections from the hippocampal formation to the thalamus were investigated with both anterograde and retrograde tracers. Horseradish peroxidase was injected into medial and midline thalamic sites in six cases, and tritiated amino acids were injected into the hippocampal formation in nine others, five of which had prior transections of the fornix. Only the subicular and entorhinal cortices were found to project to the thalamus. From the subicular cortex, dense bilateral projections were traced through the fornix to the anterior nuclei, while lighter fornical projections terminated in other rostral midline sites, including the nuclei reuniens, centralis latocellularis, and paraventricularis. These projections arose predominantly from the polymorphic cells which are located in the deepest cellular layers of the subiculum and prosubiculum. In addition, the subicular cortex was found to project to the nucleus lateralis dorsalis. The latter projection, which showed evidence of a crude topographic organization, ran either through the fornix or, unlike the other subicular efferents, through the sublenticular limb of the internal capsule to form part of the temporopulvinar bundle of Arnold. The nonfornical projection to the nucleus lateralis dorsalis passed through the medial pulvinar, where there was some additional termination. Few, if any, projections from the entorhinal cortex to the thalamus travelled in the fornix. Rather, the entorhinal efferents were carried in the inferior thalamic peduncle to the magnocellular portion of the nucleus medialis dorsalis, and in the internal capsule and bundle of Arnold to the medial pulvinar and the nucleus lateralis dorsalis.

The performance of a group of three postencephalitic subjects with anterograde amnesia was examined on a series of concurrent visual discrimination problems and on a test of visual recognition, delayed matching-to-sample. These tests were chosen as they have been used to assess experimental models of anterograde amnesia in nonhuman primates. In comparison with a group of normal subjects the postencephalitic group were impaired on the more difficult concurrent discrimination problems. They also performed poorly on the matching-to-sample task when given lists of items to remember or given increased retention intervals. The pattern of performance of the postencephalitic group matched closely that of a group of Korsakoff subjects, indicating that these behavioural tests are equally sensitive to different types of anterograde amnesia.

Thalamic contributions to memory were assessed in monkeys with lesions placed in the medial portions of either the anterior or posterior thalamus (AMT and PMT respectively). Nine cynomolgus monkeys (Macaca fascicularis) were trained on a recognition task which required the animal to select a novel object in preference to a sample object which had been presented to the monkey 10s earlier. Three monkeys then received surgical lesions centred in the anterior nuclei (AMT) and three received lesions centred in nucleus medialis dorsalis (PMT). Although the experimental monkeys rapidly relearned this task of object recognition, both the AMT and PMT groups performed significantly worse than the unoperated control monkeys when the task was made harder by requiring the animals to distinguish the sample objects after increasing delays (30s, 60s, 120s) or after additional intervening objects (3, 5, and then 10). Furthermore, all three PMT animals and two of the AMT group were moderately impaired on a test of object-reward association memory in which the animals had to remember which of two sample objects had been previously paired with a good reward. Comparison of these results with those of a previous study in which the AMT and PMT regions were removed jointly suggests that damage in either region can induce a memory loss but that combined damage to both is required to produce the full amnesia.