Library

Notes: Summary 1. Horizontal eye and head movements made by freely swimming goldfish have been recorded cinematographically and analyzed frame-by-frame. 2. Most horizontal eye movements occur during turns. A binocular saccade precedes a turn, and the eyes counterrotate relative to the head as the turn progresses, thus keeping a nearly constant orientation in space. The subsequent saccades reset the eyes from an intermediate position to an extreme one. 3. Both the saccades and the compensatory movements are generally unequal in the two eyes. The rotation by the eye on the outside of a turn exceeds that of the inner one, which undercompensates. 4. The compensation factor, CF, defined as the ratio of (mean binocular rotation relative to the head)/(head rotation), is variable from turn to turn, but averages 0.95±0.10 (mean±2 S.E.M.). 5. The slight undercompensation by the inside eye, when coupled with forward movement by the fish, results in relative visual stabilization of a region of space some tens of centimeters or less lateral to the animal. This stabilized region changes with each saccade.

Notes: Summary 1. Goldfish were deprived of visual input and/or normally functional horizontal semicircular canals. 2. Their horizontal eye movements were measured from cinematographic records, and the extent to which the eyes compensated for horizontal rotations of the head was given by the “compensation factor,” CF, the ratio (rotation of the eyes relative to the head)/(rotation of the head). All data were obtained from freely-swimming fish. 3. The CF's for normal, canal-lesioned, blinded, and blinded/canal-lesioned animals were: −0.95±0.10, −0.80±0.10, −0.50±0.04, and −0.41±0.06 (means ±2 S.E.M.), respectively. 4. These figures lead to the conclusion that the visual input contributes −0.39 to −0.45 to the CF, while the canal input contributes only −0.09 to −0.17. Thus, the visual input is the major factor; the canals are quantitatively much less important. There is in addition a third source (or sources) contributing to the CF, since the blinded/canal-lesioned fish compensated partially. 5. Experiments of partially restrained animals showed that this third source is not a preprogrammed instruction, nor is it dependent on sensory feedback from the rest of the labyrinth, tactile receptors, lateral line current detectors, or proprioceptors in the trunk. Its identity remains a mystery.

Notes: Summary The features of unlesioned and lesioned trochlear nerves of goldfish have been examined electron microscopically. Lesioned nerves were studied between 1 and 107 days after cutting or crushing the nerve. 1. Unlesioned nerves contained, on average, 77 myelinated axons and 19 unmyelinated axons. The latter were found in 1–2 fascicles per nerve. A basal lamina surrounded each myelinated axon and fascicle of unmyelinated axons. The numbers of myelinated axons, fascicles of unmyelinated axons and basal laminae varied by less than 5% over the intraorbital extramuscular segment of the nerve. 2. Following interruption of the nerve, by either cutting or crushing, all of the axons and their myelin sheaths began to degenerate by 4 days in the distal nerve-stump. Both abnormally electron-dense and electron-lucent axons were observed. Both Schwann cells and macrophages appeared to phagocytose the myelin sheaths. 3. Following a lesion, the Schwann cells and their basal laminae persisted in the distal nerve-stump. In crushed nerves, the basal laminae surrounding myelinated axons formed 97%, on average, of the Schwann tubes in the distal stump. The perimeters of the basal laminae were of similar size to those in the proximal stump, at least for the first 8 days after crush. 4. In crushed nerves, single myelinated axons in the proximal nerve-stump gave rise to multiple sprouts, some of which reached the site of crush by 2 days, the distal stump by 4 days and the superior oblique muscle by 8 days. The regeneration of the unmyelinated axons was not examined. 5. In both crushed and transected nerves, nearly all of the sprouts in the proximal and distal stumps were found within the basal laminae of Schwann cells, even though the sprouts were disorganized in the transected region where there were no basal laminae. The growth cones of the regenerating axons were always found apposed to the inner surface of the basal laminae, which may have provided an adhesive substrate that directed their growth. 6. Terminal sprouts from the ends of myelinated axons in the proximal stump accounted for the majority of the regenerating axons in the distal stump, as only a few collateral sprouts were found in the proximal stump, and only a small amount of axonal branching was found within the distal stump itself. 7. The largest axons in the distal stump were remyelinated first, and the number of remyelinated axons increased progressively between 8 and 31 days after crush, at which time there were about twice as many as in unlesioned nerves. The number of remyelinated axons remained constant at least until 107 days, the longest time considered, and none was observed to degenerate, whereas some axons that were not remyelinated appeared to degenerate. 8. Although each basal lamina in the distal stump often surrounded several regenerating axons during the first 2 weeks post-lesion, each remyelinated axon became individually surrounded by a basal lamina, collagen fibres and extracellular space between 13 and 107 days, thereby increasing the number of basal laminae in the distal stump. 9. Regenerated axonal terminals in the superior oblique were first observed 8 days after crush. The number of synapses increased progressively between 8 and 107 days, at which time they were as numerous as in unlesioned animals.

Notes: Superior oblique muscle/trochlear nerve pairs from goldfish of various ages (1-5 years) have been examined light and electron microscopically. The muscle grows by enlargement (longitudinally and transversely) of individual fibers, and by addition of new ones at the rate of about 250/year. The nerve grows by enlargement of fibers, but few and perhaps nonew axons are added. The somata enlarge, and the neuromuscular synapses become much more numerous. The ratio of muscle fibers to nerve fibers increases from about 5 in the young to about 16 in the old fish.

Notes: The selective connections that nerve cells make to each other and to effector and receptor cells outside the nervous system are essential to organized behavior. The retinotectal projection has been used frequently in the experimental investigation of how such patterned sets of connections are formed. This article traces the evolution of some of the dominant experimental paradigms and concepts that have resulted in the currently accepted rules underlying the retinotectal map.