(Continued from previous page)
The eventual recovery of boxing behavior by some animals with extensive central gray lesions, although the boxing remained below normal levels, suggests that there are cells outside the central gray which can also participate in mediating the behavior. These data from the rat are comparable to findings in the cat  that initial deficits in affective defense produced by lesions of the central gray remained for several weeks, but after that time normal behavior began to reappear. One likely explanation for the recovery of these behaviors is that cells in the reticular formation lateral to the central gray may also be able to perform the functions formerly performed by the central gray. Anatomically, there is a basis for comparing the two areas; their cells have overlapping dendritic fields, and they have been classified as part of a diffuse anatomical continuum, the so-called isodentritic core of the brain stem . Physiologically, there is also overlap between the two areas in the cat; both stimulation and recording techniques suggest the areas involved in affective defense behavior include both the central gray and also the reticular formation several millimeters lateral to it . If recovery of function is mediated by the reticular formation lateral to the central gray, one would expect lesions which destroyed both regions to produce permanent deficits; in the present series, the one animal with such extensive lesions showed such a deficit with no signs of boxing recovery after three weeks of testing.
Escape and freezing behaviors also appear to require the participation of the midbrain central gray, as had been reported by previous investigators [12, 17, 19], and was confirmed in the present study. The present results are consistent with the conclusions of Liebman, Mayer, and Liebeskind  that "the integrity of the ventrolateral central gray is necessary for the normal expression of fear." It may be significant for the behavioral relationships among defense, escape, and freezing that overlapping brain areas are important for all three. Since the three behaviors are incompatible responses evoked by similar stimulus situations, one may ask how an animal chooses which behavior to exhibit at anyone time. A rat may initially freeze in response to pain or another approaching animal, then change to flight if the approach is threatening, and finally change to defensive boxing if escape is blocked. In the usual laboratory test of boxing behavior, escape is blocked by confining the tested animals in a small enclosure; if the enclosure is made larger there will be much less boxing .
There are at least two ways that the central gray might play a role in the animal's commitment to a particular response when confronted with pain or threat. Following the theory of Kilmer and McCulloch  one might expect the central gray to function as a modal command system, gathering information about the relative advantages of freezing versus flight and flight versus turning to fight and then committing the animal to one behavior or another.
Alternatively, if one follows the theory of Flynn , one might hypothesize that the central gray functions as a patterning mechanism, establishing a readiness for action in several patterning reflexes such as a reflex to freeze, a reflex to run towards a safe place of an escape route, and a reflex to stand upright and box if another animal rises in front of it. Depending upon the environmental stimuli, one or another (or a combination?) of these patterned reflexes would be carried out.