2. Lesions thought to be characteristic of poliomyelitis in inoculated chimpanzees could not be sharply differentiated from lesions of unknown origin in uninoculated control animals. Moreover, although the inoculated animals as a group, in comparison with the control animals, had a greater number of infiltrative lesions in sympathetic as well as in sensory ganglia, it was not possible to make satisfactory correlations between the distribution of these lesions and the routes of inoculation.

3. In sharp contrast with chimpanzees, the celiac and stellate ganglia of the human poliomyelitis cases were free of any but insignificant infiltrative lesions. Lesions in human trigeminal and spinal sensory ganglia included neuronal damage as well as focal and perivascular inflitrative lesions, as is well known. In most ganglia, as in monkey and chimpanzee sensory ganglia, these were correlated in intensify with the degree of severity of lesions in the region of the CNS receiving their axons. This suggested that lesions in sensory ganglia probably resulted from spread of virus centrifugally from the CNS, in accord with considerable experimental evidence.

4. Two principal difficulties in the interpretation of histopathological findings in peripheral ganglia were revealed by this study. The first is that the specificity of lesions in sympathetic ganglia has not been established beyond doubt as being due to poliomyelitis. The second is that the presence of characteristic lesions in sensory ganglia does not, and cannot, reveal whether the virus reached the ganglia from the periphery or from the central nervous system, except in very early preparalytic stages or in exceptional cases of early arrest of virus spread and of lesion production.