ATMOSPHERIC TIDES AND OSCILLATIONS 
world.® Martyn finds at latitudes 35°S and 27°S a lunar 
variation in the height of the H-region which is opposite 
in phase to that found by Appleton and Weekes in the 
higher latitude of England. Martyn also finds lunar 
variations in both the heights and electron densities of 
the F-region. Near the magnetic equator these varia- 
tions are very much larger than the already large vari- 
ations found in the H-region. At Huancayo (Peru) the 
total tidal variation at certain hours and seasons 
amounts to some 60 km in height and 20 per cent in 
electron density. These variations concern the layers of 
electrons and ions interspersed among the neutral mole- 
cules of the high atmosphere. Their relation to the tidal 
oscillations of the main body of air requires considera- 
tion of electrodynamic as well as of hydrodynamic 
forces. There is no doubt that when the theory of these 
variations is fully understood the observational results 
will provide important and interesting information 
about the lunar atmospheric tide at high levels. 
Fig. 14.—Harmonice dial [6] for the lunar tide in the H-region 
of the ionosphere. The circle shows the probable error for any 
one of the eleven separate dial points, each determined from 
12-14 days’ data. 
Lunar Tidal Variations of Cosmic Rays. Cosmic ray 
observations provide, in a surprising and most interest- 
ing way, information as to the lunar air tide at a level 
of eighteen or twenty kilometres above the ground, 
though a precise interpretation of the data awaits fur- 
ther study. Among the cosmic rays received at the 
ground are mesons, supposed to be generated (by the 
primary rays) at this level; being unstable, a propor- 
tion of them are transformed on their way to the ground. 
If the lunar tide raises or lowers the level of the mean 
air pressure at which the mesons are generated, their 
path to the ground will be lengthened or shortened, and 
the number of survivors at ground level will be reduced 
or increased. Duperier [59] has made a reliable deter- 
mination of LZ» in his recorded amounts of cosmic ray 
reception (mainly of mesons) at London, and has in- 
ferred therefrom that the lunar tide at about 18 km 
6. For other lunar ionospheric tidal determinations see 
[4, 5, 26, 81], and a forthcoming report by D. F. Martyn in the 
Ziirich (1950) Proceedings of the International Union for Scien- 
tific Radio (U.R.S.1.). The lunar diurnal variation in the 
thickness of the F3-layer in Alaska, reported by M. W. and 
J. G. Jones, in J. Meteor, 7: 14-20 (1950), is not real. 
521 
height is considerably magnified (about tenfold as com- 
pared with the equilibrium tide), though much less than 
in the ionospheric E-layer. 
The Geomagnetic Lunar Tide. Nearly a century ago 
a small lunar daily variation was detected in the rec- 
ords of the components (or ‘“‘elements”) of the earth’s 
magnetic field. This variation is produced (lke the cor- 
responding solar daily geomagnetic variation) mainly 
above the earth’s surface, though these varying “pri- 
mary” magnetic fields of external origin induce electric 
currents in the conducting body of the earth—mainly 
deep down, but also, to a lesser degree, near the surface, 
where they can be measured and recorded. The analysis 
of these earth-current records reveals solar and lunar 
daily variations, which form yet another curious by- 
product, as in the cosmic rays, of the high-level solar 
and lunar tidal atmospheric oscillations. 
The external source of these solar and lunar daily 
geomagnetic variations consists of systems of electric 
currents flowing in some layer or layers (not yet clearly 
identified) in the ionosphere, and these currents are in- 
duced by mainly horizontal oscillatory large-scale mo- 
tions of the ionospheric air. The process is similar to 
that in a dynamo, as first suggested by Balfour Stewart 
[42, 49]. The moving air corresponds to the armature, 
the conducting ionospheric layers to the armature wind- 
ings, and the earth’s main magnetic field to the field of 
the dynamo pole pieces. 
From the determinations of S, and L, in the mag- 
netic records of many stations it is possible to determine 
the distribution and intensity of these solar and lunar 
daily-varying electric current systems in the ionosphere, 
and also the type of the inducing atmospheric motions 
at those levels. To infer the intensity of these motions 
requires a knowledge of the electric conductivity of the 
layers in which the known electric currents flow. Until 
the precise situation of the currents is ascertained, and 
their electric conductivity, the intensity of the solar 
daily and lunar daily oscillations in the ionosphere can- 
not be precisely inferred from the geomagnetic data. 
The present indication is that the lunar tidal horizontal 
movements, like the lunar tidal rise and fall of the E- 
layer, are very greatly magnified as compared with what 
the barometric L. data would suggest. It is, however, 
of interest to note that the ratio of S. to Ly in the mag- 
netic records is about the same as that in the barometric 
variations. There is much scope for further investiga- 
tion, both by observation and theory, of the bearing 
of the geomagnetic data on the solar and lunar daily 
atmospheric oscillations. 
THE SOLAR SEMIDIURNAL OSCILLATION S, 
The Components S, of the Solar Daily Barometric 
Variations. In middle and high latitudes the barometric 
variations are large, and mainly connected with weather 
changes. By averaging over many days selected in any 
way—days of a given season or calendar month, or 
days of high barometer or of rain—characteristic daily 
barometric variations can be determined. In the tropics, 
where large irregular barometric changes are rare, S2 
