New evidence for a latitudinal decrease of the solar wind mass flux is presented from observations of the interplanetary Lyman alpha emission collected in 1976 and 1977 with satellites Prognoz 5 and 6. The flow of interstellar hydrogen atoms in the solar system is ionized by EUV solar radiation and charge exchange with solar wind protons which accounts for about 80% of the total ionization rate. The resulting gradual decrease of the neutral H density from the upwind region down to the downwind region observed from Ly &agr; intensity measurements allowed the determination of the absolute value of the total ionization rate &bgr; for one H atom at 1 AU against ionization. Collected in 1976 and 1977 at five places in the solar system, The measurements are first compared to a model which assumes isotropy of the EUV and solar wind. Strong departures are obvious toward high-latitude regions, especially when the observer is in the downwind region where the solar wind ionization has had more time to act (cumulative effect). A model was constructed which include a decrease of the ionization rate with heliographic latitude. The adjustment of data allowed for the measurement of the absolute value of the total ionization rate and implies a 50% latitude decrease of the ionization rate due to charge exchange with the solar wind, from &bgr;sw=(3.9¿0.5)¿10-8 s-1 at the equator to &bgr;sw=(2.0¿0.5)¿10-8 s-1 at the pole. The corresponding absolute value of the solar wind proton flux is (2.4-3.6)¿108 cm-2 s-1 at the equator and twice less at the pole if a constant velocity is assumed for the solar wind. Even if the solar wind velocity increases from 400 to 800 km s-1, which would decrease the charge exchange cross section by 25%, there is still a decrease by about 30% of the solar wind mass flux from equator to pole. The Lyman alpha data from Mariner 10 [Kumar and Broadfoot, 1978> had already shown a similar trend in 1974, showing a persistence of the solar wind anisotropy for 3--4 years during a solar minimum. The large-scale properties of the solar wind mass flux can therefore be monitored at all latitudes by remote sensing of Ly &agr; interplanetary emission, since the solar wind is carving the flow of interstellar H and its anisotropies are ''printed'' on the interplanetary H distribution. With uncalibrated Ly &agr; measurements, an absolute value of the solar wind flux can be determined at all latitudes averaged over a typical 1-year period.