Figure 6. Phase tuning of fast, large-amplitude EPSPs resembles the phase modulation of the average membrane potential. A, Polar plots of the vector magnitude by the vector angle for all 18 GCs with significant membrane potential respiratory coupling. Vectors calculated from the average membrane potential in each GC of all EPSP (gray, left), fast, large EPSP (F/L EPSPs; orange, middle), and slow, small EPSP (S/S EPSPs; green, right) subpopulations. B, Plot of the mean vector magnitudes calculated the mean membrane potential (black), from all EPSPs (dark gray), from F/L EPSPs (orange), and from S/S EPSPs (green). **p < 0.005, ***p < 10−7. C, The average absolute angular difference between the membrane potential vector and the vector calculated from different group of EPSPs. All EPSPs (gray) and F/L EPSPs (red) are both significantly closer to the membrane potential vector than S/S EPSPs (*p < 0.05 and **p < 0.005, respectively). D, Plot of absolute angular difference between the membrane potential vector and the F/L EPSP vector magnitude for GCs with F/L EPSP vector distributions significantly different from uniform distributions (p < 0.05, Rayleigh's test, n = 15 of 18 GCs). A dashed line shows a trend in the data that are not significant (r2 = 0.030, p = 0.25). E, Plot of the absolute angular difference between the membrane potential vector and the S/S EPSP vector magnitude, including only cells with S/S EPSP vector distributions significantly different from uniform distributions (p < 0.05, Rayleigh's test, n = 13 of 18 GCs). A dashed line illustrates a significant, negative correlation between S/S EPSP vector magnitudes and the difference between the angle of the S/S EPSP vector and the membrane potential vector (r2 = 0.42, p < 0.01).