N = North geographic pole Figure 3. Earth's magnetic field. The broken lines depict the lines parallel to the direction of magnetic force. As became increasingly clear over the years, the actual magnetic field of the earth differs considerably from this idealized picture of a dipole field. Chapman, Ferraro, and others. Then the gradual recovery from this "main phase" of the magnetic storm, as it was called, signified the gradual dissipation of the ring current and a return to normal conditions-or so it was thought. Among the most notable of high-altitude phenomena, and among the earliest to be studied in detail, were the auroras, the northern and southern lights. These were seen most frequently at heights from 90 to 120 km, but also occurred at both lower and much greater heights. That the auroras correlated strongly with activity on the sun and appeared in an auroral belt at high latitude suggested that they must be due to charged particles from the sun. Charged particles would be steered by the earth's magnetic field, whereas neutral particles or solar photons would not be affected by the earth's field. Experimenting with cathode rays and small magnetized spheres, K. Birkeland in 1898 and others demonstrated how electrified particles approaching a magnetized sphere from a distance would be guided by the magnetic field toward the poles. Starting from Birkeland's concepts and experiments, over many years Carl Størmer developed a theory of how electrons or protons from the sun would be deflected by the earth's magnetic field into the auroral zones to produce the auroras as the particles impacted on the atmospheric molecules, causing them to glow.16 The spectrum of the aurora was observed to exhibit primarily lines and bands of atomic oxygen and molecular nitrogen, with the forbidden green lines of atomic oxygen at 5577 Å being particularly strong. At nighttime the high atmosphere was seen to emit a very faint light, sometimes called the permanent aurora, also consisting of the forbidden lines of atomic oxygen and of bands of the nitrogen molecule. This air glow was estimated to come from well above 100 km, perhaps from as high as 400 to 500 km, very likely from F-region ions as they were neutralized during the night. The yellow sodium D lines were also seen emanating from the lower part of the E region, and were particularly intense at twilight. From a distant cloud of material particles of some sort, the zodiacal light, with a spectrum similar to that of the sun, contributed to the light of the night sky. In the mid-1940s it was not known whether this radiation came from within the high atmosphere or from interplanetary space. At some height, probably around 800 or 1000 km, the atmosphere was expected to cease acting like a normal gas. In this region collisions between atmospheric particles would be infrequent, and a molecule might rise along an elliptic orbit to an apogee and fall back without colliding with another molecule until returning to the denser atmosphere at lower altitudes. If the molecule had sufficient velocity it might even escape into interplanetary space. Indeed, it was supposed that hydrogen and helium had to be escaping continuously through this fringe region, even though neither had been detected in the upper atmosphere. Helium was known to be entering the atmosphere from the ground-where it was produced by the decay of radioactive elements-at a small but measurable rate; but the percentage of helium in the lower atmosphere remained constant over time. The natural conclusion was that this light gas had to be diffusing up through the atmosphere to the highest levels where the very high temperature permitted a ready escape of the gas. Somewhere in this fringe region, or exosphere, the transition from the earth's atmosphere to the medium of interplanetary space was assumed to lie. One was hard put to it to define the boundary. Presumably where the atmospheric density had dropped to the few particles per cubic centimeter expected in interplanetary space the boundary must already have been crossed. But long before then the atmosphere had ceased to exist in the usual sense of the term. Across this ill-defined interface, radiations from the sun entered the earth's environs to cause the auroras, magnetic storms, ionization, and heating of the atmosphere. Across this interface also came the cosmic rays. 17 These highly energetic particles from outer space were more the concern of the high-energy physicist than of the geophysicist. Discovered between 1911 and 1914 from balloon experiments on atmospheric ionization, cosmic rays quickly became a subject of intense interest. It was soon accepted that the rays came from outside the earth. Measurements of the ionizing power of the rays at various depths below the surfaces of mountain lakes revealed both a soft component and a hard, or extremely penetrating, component to the rays. Balloon experiments showed that the intensity of the radiation increased steadily with altitude until a maximum-called the Pfotzer maximum— was reached at about 20 km in mid-lititudes. The shape of these intensityaltitude curves is shown in figure 4a.18 Figure 4b shows schematically that Figure 4(a), at left. Cosmic rays at high geomagnetic latitudes. Figure 4(b), at right. Geomagnetic effect on cosmic rays. A schematic drawing showing how, according to experimental measurements, cosmic ray intensities vary with geomagnetic latitude. See Bowen, Millikan, and Neher in Physical Review 53 (1938): 855-61. the earth's magnetic field has a distinct effect upon the radiation, leading to the conclusion that the rays are charged particles, not photons. The shape of the intensity-altitude curve was explained as follows. The primary rays, whatever they might be, upon striking the atmosphere produced a shower of secondary rays, which, added to the primary rays, caused the initial increase in total ionization observed at high altitude. Eventually, however, an equilibrium was reached, with the atmosphere absorbing enough energy from both the primary and secondary particles to decrease the total ionizing power with further depth into the atmosphere. Such a transition curve, as it was called, would be observed not only in air, but also in lead or other substances, the principal difference being the spatial extent of the transitions, which depended on the density and nature of the material. The early idea that the primary cosmic rays might be high-energy electrons was soon rejected. It could be shown that to penetrate the entire atmosphere and reach the ground, electron showers would have to be caused by primary electrons with such high energy that they would be completely unhindered by the earth's magnetic field. They would accord ingly not exhibit the magnetic field effect already shown to exist. In 1938 T. H. Johnson concluded that the primary radiation consisted of protons, as theorists had guessed somewhat earlier. In 1941 balloon observations revealed that the cosmic rays within the atmosphere at high altitude were mostly mesotrons (mesons), presumably generated by the primary protons. 19 No significant component of electrons was observed at high altitude, supporting the conclusion that there could be no significant component of electrons in the primary radiation. But the soft component observed near the ground was believed to be electrons, decay products of the mesons generated at high altitude. PROBLEMS TO SOLVE Thus, the scientific paradigm for the earth's upper atmosphere in the mid-1940s was rich in ideas accumulated over more than half a century of observation and theoretical study. It had been possible to explain to a considerable degree a wide range of phenomena, many of which proved to be extraordinarily complex; but many uncertainties, unanswered questions, and problems remained. Consider the problem of estimating atmospheric densities in the E region of the ionosphere around 100-km altitude. In the 1920s F. A. Lindemann and G. M. B. Dobson approached this problem by using observational data on the heights of appearance and disappearance of visual meteors. Intuitively it seemed reasonable that the density of the gas traversed by a speeding meteor should play a role in determining where the meteor would glow and be visible. The challenge was to develop a suitable theory to relate the observed meteor trails to the atmospheric density. Lindemann and Dobson assumed that as the meteor rushed into the atmosphere, a hot gas cap formed because of compression of the air. Heat from the gas cap was transferred to the meteor, and if the object were small enough it became incandescent. Making a number of assumptions about how heat was transferred from the gas cap to the meteor and using kinetic theory, Lindemann and Dobson derived expressions for p, the density of the atmosphere at the height of appearance, and p, the density at the height of disappearance of the meteor. The equations are reproduced here to emphasize the large number of quantities involved, uncertainties in which could cause errors in the derived atmospheric densities. Mo k V1 = (V1-V2)/3v = calculated efficiency factor of heating = velocity of the compressed gas molecules in front of the meteor V2 = velocity of the gas molecules at the temperature of the meteoric V, 2 surface l = latent heat of vaporization of meteoric material υ R То = = velocity of the meteor, assumed constant = universal gas constant = To temperature of the atmosphere, assumed isothermal throughout the range of consideration L = total length of the meteor trail From the apparent brightness of the meteor the rate at which energy was being emitted could be calculated, which multiplied by the time of visibility gave the total amount of energy radiated. Setting this equal to the kinetic energy 1⁄2mv2 yielded the mass m of the meteor. If one then assumed that the meteor was iron and essentially spherical, one got from the expression mass density times volume = m = pm. (4πr3/3) (6) which gave the radius r. The other quantities in the expressions for the atmospheric density could be either measured directly or estimated from plausible assumptions, thereby giving densities at two altitudes, that of appearance and that of disappearance. The chain of reasoning was lengthy, with many assumptions. The results obtained by the investigators immediately put some of the assumptions into question. For example, the air densities obtained proved three times too high to correspond to an isothermal atmosphere at the stratospheric temperature of 220 K, requiring instead temperatures around 300 K. Between the stratosphere and the E region of the ionosphere, then, there had to be a significant variation in temperature. Moreover, other observa |