We determined the phases of the maxima (spring, fall) and minima (summer, winter) in the curve of smoothed daily averages of the aa geomagnetic index, available from 1868 to 1998. The dates we obtained are consistent with the equinoctial hypothesis which has aberration-adjusted theoretical maxima, for a ~440kms - 1 (modern epoch) average solar wind speed, on 25 March (experimentally determined to be 27 March, with an uncertainty of +/-2 days) and 27 September (27 September) and minima on 25 June (26 June) and 26 December (27 December). We also show that the overall shape of the 30-day smoothed modulation curve throughout the year (broad minima, narrow peaks) bears greater fidelity (|r|=0.96) to the aberration-shifted solar declination δ (the controlling angle, on average, for the seasonal variation under the equinoctial hypothesis) than to the solar B 0 angle (r=0.76; axial hypothesis) or the solar P angle (r=0.86; Russell-McPherron effect). Lastly, a three-parameter fit of the smoothed annual variation of the aa data with a function consisting of the sum of the smoothed yearly curves for the δ, B 0 , and P angles yielded an amplitude of 0.58+/-0.07 for the δ component vs. 0.16+/-0.03 for B 0 and 0.20+/-0.04 for P. Thus, the phases and profiles of the 6-month wave in the long-running mid-latitude aa range index are consistent with control by a dominant equinoctial mechanism.