RESPONSE OF ANTENNAL COLD RECEPTOR OF PERIPLANETA AMERICANA TO RAPID TEMPERATURE CHANGES AND TO STEADY TEMPERATURE

1. Ventrally located cold sensilla on the antennae of male adult Periplaneta americana are ring-shaped structures about 8 μ in diameter with a short hair emerging from their center (Figs. 1, 2). 2. Further experiments (cf. Loftus, 1966) have substantiated the exclusion of the motion of air streams (2.5 m/sec) as a stimulus. The list of pure and mixed chemicals has been extended. The results of all tests including the odor of crushed abdomens of virginal female cockroaches were negative. 3. Extracellularly recorded impulses which were either monophasic negative or diphasic at higher temperatures often became either diphasic or monophasic positive at lower temperatures. The effect was reversible (Fig. 4). 4. When peak frequency is driven over 200 imp/sec, the time course of instantaneous frequency, F=f(t), plotted on double log scale displays 3 linear phases: 1 rising and 2 falling, followed by a fourth phase of recovery. At lower peak frequency only one falling phase develops. Before the onset of the recovery phase instantaneous frequency values determined from single impulse intervals lie within ±20% of a linear function on double log scale (Figs. 6, 7, 8). 5. Instantaneous frequency values, even when formed from groups of as many as 10 successive impulse intervals, cease to convey unambiguous information on stimulus magnitude within about 300 msec after peak frequency (Figs. 9, 10). 6. The reciprocal of the total number of impulses from stimulus onset up to about 800 msec afterwards (mean frequency), shows a much more precise relationship to ({ΔT) than instantaneous frequencies after Fmax do (Figs. 12, 13). 7. When instantaneous frequency is plotted against temperature drop, the slope of the curves, F=f({ΔT), falls off with increasing ({ΔT) and time since stimulus onset. Thus the sensitivity of peak frequency is the highest of all instantaneous frequencies and itself higher at smaller temperature drops. An average receptor attains a sensitivity of -155 imp/sec/°C (Fig. 11). 8. When peak frequency is plotted against initial temperature and temperature drop, a bulging surface develops. Hence a given peak frequency determines only a set of values for initial temperature and temperature drop. Three receptors with different characteristic surfaces should permit both components to be distinguished quantitatively (Fig. 14). 9. As shown in peak frequency, the receptor's sensitivity to temperature drop diminishes, whereas the effect of initial temperature on peak frequency rises, towards the extremes of the temperature range investigated (Fig. 15). 10. Treatment of 13 receptors and 280 responses indicate sufficient reliability of peak frequency response to permit discrimination of 7 to 17 discrete temperature drops between ({ΔT)=0.2° and ({ΔT) = 5.5° C (Fig. 16). 11. A second identical short temperature drop repeated after 1.5 sec rewarniing evokes a peak frequency on the average 15% lower than the first. This effect could diminish resolving power severely (Fig. 17). 12. The longest single impulse intervals occur at rapid temperature rises. Reckoned as frequencies they attain values in the order of 0.1 imp/sec. These frequency minima show good quantitative relationship to the extent of temperature rise and indicate the possibility of this cold receptor's providing the CNS with information on warming (Fig. 18). 13. Some receptors show a sharp activity maximum between 17° and 37° C in response to steady temperature. The responses (mean frequencies) of many receptors, however, manifest variation up to and beyond 20% of the frequency range corresponding to steady temperature between 17° and 37° C. Hence the receptor is a poor thermometer (Figs. 19, 20, 21). 14. Taken absolutely the sensitivity of Periplaneta's cold receptor to temperature change (-155 imp/sec/°C) seems second only to infrared sensitive crotalid facial pits (+500 imp/sec/°C). © 1968 Springer-Verlag.