(Continued from previous page)
Under circumstances of facial tactile inactivation, there is an alternative visual system, including the lateral geniculate and visual cortex, which can release the behavior. It does not direct the behavior, however, as indicated by the fact that tracking of the opponent does not occur. Somehow, this thalamo-cortical visual system is capable of utilizing the thalamo-cortical projections of the tactile system to re-establish the functions previously carried out by lower levels of the tactile system. This can explain why ventrobasal lesions, as studied in this paper, can block the assumption of visually-released boxing. It may also explain why lesions of the medial lemniscus also block visually-released boxing, if one assumes that medial lemniscus inputs to the ventrobasal thalamus are essential for its proper functioning.
A number of questions remain concerning the neural circuitry of the defensive upright posture and boxing. What is the neural relationship between the visual cortex and ventrobasal thalamus? Are they related by way of the striatum, sensory cortex, or other intervening structures in the mediation of visually-related boxing? Also, where is the motor patterning mechanism of the upright posture and boxing, and how does it receive releasing and directing inputs from the visual system during visually-released boxing?
The present study confirms the usefulness of a methodology in which the brain pathways of specific behaviors are traced by lesions in animals which have been forced to rely upon one sensory system to release and direct these specific behaviors. It is well known that mammals can recover function from many kinds of brain lesions by employing alternative sensory systems. By depriving the animal of all relevant sensory systems except one, however, it is no longer possible for the animal to compensate in this fashion, and the effectiveness of specific lesions can be assessed without ambiguity.
Visually-released boxing is a behavior which may prove fruitful for future experiments in a number of areas of traditional interest. Not only is it proving useful for studies in brain research, but also it promises to be of interest for studies of behavioral plasticity and cross-modal transfer. As pointed out by Thor and Ghiselli  and confirmed in our laboratory , naive rats are not able to adopt visually-released boxing after facial anesthesia, but only experienced rats can do so. What is the nature of the learning process which distinguishes these two groups of animals? Since the thalamus and cortex appear to be intimately involved in the neural circuitry of this learning or cross-modal transfer, visually-released boxing may provide a useful model to study the role of the cortex in these processes, a role which has long been hypothesized but rarely documented in specific detail.