Pc3-4 (15--50 mHz) pulsations are often observed during daytime hours in the Earth's magnetosphere and on the ground. The ion foreshock region of the solar wind upstream of Earth is now understood to be a major source of these waves, and the general dependence of their occurrence in the magnetosphere on the cone angle of the interplanetary magnetic field (IMF) is well attested. However, there have also been suggestions that the IMF clock angle should introduce latitudinal or local time asymmetries in wave occurrence, and in particular, that Pc3-4 pulsations might preferentially reach northern versus southern high-latitude regions. In order to test this latter suggestion, we have compared a full year's (1997) data from two cusp-latitude stations, Sondrestromfjord, Greenland, and South Pole, Antarctica, both near 74¿ corrected geomagnetic latitude. Our data show the usual strong IMF cone angle control of Pc3-4 wave occurrence/amplitude but also indicate that occurrence and relative amplitude are essentially identical near both cusps; any clock angle control is at most a second-order effect, not detectable in our analysis. We further find little seasonal dependence in the ratio of power in the two cusp regions other than that attributable to differences in solar illumination. These observations, in conjunction with the well-known sharp cutoff in wave activity for cone angles above 45¿, indicate that only those perturbations occurring at the subsolar bow shock can reach the magnetopause and propagate to any location in the dayside magnetosphere. In addition, the absence of a clock angle effect supports earlier suggestions that Pc3-4 signals in the magnetosheath do not cross streamlines and thus do not propagate as waves, but are rather convected as spatial structures in the highly turbulent, high-beta downstream magnetosheath plasma.