Measurements of the cross polar cap electric potential, by the double probe electric field experiment aboard S3-3, from 55 orbits in the dawn-dusk plane are compared with the reconnection electric fields predicted by a variety of models, both theoretical and experimental. The purpose of these comparisons is to understand the extent to which nonreconnection contributes to the polar cap potential must be included, to determine the time response of the polar cap potential to time varying reconnection rates, and to determine the efficiency and saturation levels of the reconnection process. It is found that (1) After several hours of northward interplanetary magnetic field, the cross polar cap potential declines to progressively lower values than those after 1 hour of northward interplanetary magnetic field. This suggests that it requires several hours for the ionospheric polar cap potential to respond to the ''turning off'' of ''turning down'' of the reconnection process. (2) The decay of the polar cap potential is used to demonstrate that contirubtions to the polar cap potential not associated with the reconnection process can be limited to less than 20 kV. It is shown that contributions to the polar cap potential that scale with the dynamic pressure of the solar wind are limited to less than 1 kV. (3) The cross polar cap electric potential is best predicted by a weighted sum of contributions from interplanetary magnetic field parameter over the 4 hours previous to the measurement. The weighting functions have the form of an exponential decay 2--3 hours with the strongest weight on interplanetary parameters over the 1 hour previous to the measurement. (4) For values of the dawn dusk component of the interplanetary electric field less than about 0.5 mV/m, the measured polar cap potential is consistent with reconnection of all the interplanetary magnetic flux incident on a 30 RE wide frontside magnetopause. For larger values of the dawn dusk component of the interplanetary electric field, the proportion of field lines reconnecting is less, indicating saturation of the reconnection process. The above results are obtained when the extended response time of the magnetosphere to changes in interplanetary parameters is considered. They are independent of the detailed reconnection model assumed, and they are quite different from the visous contribution, saturation levels of reconnection, and reconnection efficiencies inferred from comparisons of polar cap potentials of single hour averages of interplanetary parameters.