ANALYSIS OF REFLEX VARIABILITY IN TERMS OF PARTIALLY CORRELATED EXCITABILITY FLUCTUATION IN A POPULATION OF MOTONEURONS

Nerve cells characteristically exhibit fluctuation in excitability. In the case of neurons lying within the central nervous system it is reasonable to suppose that multiple influences contribute to excitability fluctuation. Certain of these influences might depend upon factors intrinsic to individual nerve cells while others might be due to extrinsic factors. I t may be assumed that, among a group of neurons, coincidence of excitability swings due to intrinsic sources of fluctuation would be determined purely by chance. On the other hand, extrinsic factors might be expected to influence members of a population to some degree in unison; i.e., in a correlated manner. These considerations suggest that members of a functionally homogeneous population of nerve cells might be subject to excitability fluctuations that are in part independent and in part correlated. Such appears to occur in spinal motoneurons (3). For a fuller understanding of the reflex behavior of populations of neurons it would be desirable to determine the role of independent and correlated excitability fluctuations. Monosynaptic reflex response of a motoneuron population varies in magnitude on successive application of standard test volleys, indicating that the number of motoneurons responding varies on successive trials. A corollary of this variation in population response is that certain of its members discharge on some but not all of a series of trials. By analysis of individual motoneuron response Lloyd and McIntyre have shown that in response to a series of standard monosynaptic excitatory volleys some members of a motoneuron population never respond, others invariably respond, and a certain number respond on an intermediate number of trials (5). They employed firing index as a measure of individual motoneuron response, this index being defined as the per cent of monosynaptic reflex responses in a given number of trials. I t was found that units with firing indices at the upper and lower extremes of the intermediate range (between zero and 100) occurred more frequently than those in the middle part of this range; that is, the distribution of units with respect to firing indices was U-shaped. This was considered to result from a uniform distribution of