For both hemispheres in the neighborhood of latitude 35° the behavior of region F2 of the ionosphere has been examined. It has been found that:
(1) There are great differences in the graphs of monthly mean noon values of maximum electron‐density (
) for Washington, Wuchang, and Tokio in the north, but only small differences for Watheroo, Mount Stromlo, Wellington, and Christchurch in the south.
(2) For all stations for which data are available, midnight
is in step with the Sun, being a maximum in summer and a minimum in winter.
(3) Noon values of
correlate with values of
cos½
(where
is the monthly mean Wolf sunspot‐number, and
is the local zenith‐angle of the Sun at noon on the 15th day of the month) to give correlation‐coefficients (ρ) equal to 0.50, 0.60, and 0.25 for Mount Stromlo, Watheroo, and Washington, respectively.
(5) Midnight values of the annual mean maximum electron‐density (
) for Washington are linearly related to annual mean sunspot‐numbers (
). This follows from the fact that noon values of
bear a linear relationship to midnight values of
. In 1937 about 90 per cent of
for Washington was due to sunspots.
(6) The noon values of monthly mean equivalent heights show very great seasonal fluctuations. The midnight values are comparatively steady.
(7) In the Northern Hemisphere the graph of
cos½
for the period 1934‐40 is, for the most part, in step with the Sun, whereas in the Southern Hemisphere it shows no relation to solar altitude.
It has been concluded that:
(a) There is a variation, in the composition of region F2, from place to place over the Earth's surface.
(b) The annual effect found by Berkner and Wells may be due to the variation in composition of region F2, and to the differences in
cos½
in the two hemispheres.
(c) The annual effect has not a sidereal cause.
(d) As far as annual mean values of maximum electron‐density are concerned, the transition from noon to midnight is not influenced by the number of sunspots.
(e) If the thermal‐expansion hypothesis is correct, midnight temperatures are probably low, and do not vary much throughout the year. Summer‐noon temperatures are very much higher than temperatures at midnight.