The conduction velocities of peripheral nerve fibres conveying sensations of warming and cooling.

cleidomastoids may still continue intermittently in both forms if the patient is carefully observed. Our patient had symptoms compatible with paralytic rabies. However, the initial presentation with seizures resulted in delay in diagnosis. Clinical seizures of cortical origin were extremely rare in our experience of more than 40 rabies patients (by Hemachudha and Manutsathit, unpublished data). Only one patient who received intensive respiratory support and had severe hyponatremia had a seizure during the preterminal phase. Opistothonos or convulsions during hydrophobic spasms have been described,9 but not as an initial presentation. Attempts to make a diagnosis of human rabies in life by immunofluorescent testing for rabies virus antigen in corneal or salivary smears, or from nuchal skin or brain biopsy, and efforts to detect antibody to rabies virus in the serum or spinal fluid, were all disappointing during the early clinical stage of the disease.6 Antibody to rabies virus by RFFIT was demonstrated in this case 9 days after the onset of disease. The presence of rabies antigen of comparable amount in both neurons and glial cells was surprising. Study of rabies virus distribution in six other patients (four encephalitic and two paralytic cases) showed neuron to be the almost exclusive target of infection (Manuscript submitted). Inclusion bodies in astrocytes have been regarded as an uncommon finding on light microscopy.' " Only 17% ofhuman rabies cases reported by Tangchai et al' were found to have inclusion body positive astrocytes. These were mostly in the floor of the third ventricle, paraventricular area and brainstem. It is not known whether seizure activity can be induced by the presence of virus in glial cells. T HEMACHUDHA