Quantitative models are developed for representing the global distribution of the average magnetic field produced by the region 1 and 2 Birkeland current systems. The problem is solved in four following steps: (1) constructing a realistic tilt‐dependent model of the Birkeland current sheets, based on the formalism of Euler potentials, (2) numerically computing their field at a large number of points within the modeling region, (3) finding a best‐fit analytical approximation for that field, and (4) adding a current‐free shielding field which confines the Birkeland field within the model magnetopause. At low altitudes, the model field‐aligned currents reach the ionosphere along eccentric ovals, which fit the observed region 1 and 2 zones of Iijima and Potemra, and they continue there as horizontal currents. At larger distances, the nightside region 1 currents map to the plasma sheet boundary layer and are then diverted toward the tail flanks, while currents in the dawn‐dusk and dayside sectors connect directly to the higher‐latitude magnetopause. The region 2 current closes azimuthally near the equator, forming a spread‐out partial ring current system. The described approach allows a great flexibility in the geometry of the Birkeland currents, making it feasible to infer their properties from spacecraft data.

In a time‐independent model of the magnetosphere including a dipole magnetic field, a convection electric field (deduced from polar orbit observations), and an electric field due to the earth's rotation the motion of a proton in the equatorial plane is analyzed. By means of the Hamiltonian W of the motion, three regimes of motion are identified. A method is then developed for determining for any given observation point which arrival energies correspond to trapped particles and which represent ‘open’ trajectories.

The representation of a magnetic field by the cross product of the gradients of two scalars has recently seen wide use in plasma physics and in the study of energetic particles in space. The properties of such a representation are reviewed and examples of its application are given.

Recently, Svalgaard and Heppner reported two separate features of the polar electromagnetic field that correlate with the dawn‐dusk component Biy of the interplanetary magnetic field. This work attempts to explain these findings in terms of properties of the open magnetosphere. The topology and qualitative properties of the open magnetosphere are first studied by means of a simple model, consisting of a dipole in a constant field. Many such properties are found to depend on the separation line, a curve connecting neutral points and separating different field line regimes. In the simple model it turns out that the electric field in the central polar cap tends to point from dawn to dusk for a wide variety of external fields, but, near the boundary of the polar cap, electric equipotentials are deformed into crescents, which may explain the magnetic variation correlated with Biy and reported by Svalgaard. Modifications expected to occur in the actual magnetosphere are also investigated. The place of merging and convection in the model is outlined, and it is shown how the ‘spilling out’ of open polar field lines from a long narrow ‘window’ along the geomagnetic tail may explain the electric field asymmetries observed above the polar caps by Heppner.

To determine the effects of estrogen deficiency and replacement on GH secretion, we measured the 22-h GH secretory pattern and response to 1 h of light exercise in 16 normal postmenopausal women before and after treatment replacement with ethinyl estradiol (20 micrograms/day for 15 days). To determine whether the changes found were due to pituitary sensitization by estrogen, the response to synthetic GH-releasing hormone (GHRH; 1.0 microgram/kg, iv) was measured. To assess the biological effectiveness of GH in estrogen-treated women, somatomedin-C (Sm-C) responses to GHRH were measured. Pre- and postestrogen GH secretion rates, expressed as mean areas circumscribed by plasma GH values, were as follows: 22-h study, 1.4 +/- 0.1 (+/- SEM) vs. 2.0 +/- 0.3 ng/ml X h (P = 0.04; n = 5); during 1 h of exercise, 2.3 +/- 0.4 vs. 3.2 +/- 0.4 ng/ml X h (P = 0.03; n = 16); after GHRH-(1-40), 6.7 +/- 1.7 vs. 8.5 +/- 1.5 ng/ml X h (P = 0.12; n = 16). There also was a modest but significant increase in resting plasma GH (1.5 +/- 0.2 vs. 2.3 +/- 0.5 ng/ml (P = 0.039). Pre- and postestrogen plasma Sm-C concentrations were 0.56 +/- 0.08 and 0.32 +/- 0.03 U/ml, respectively (P = 0.006; n = 16). Thus, estrogen therapy increased spontaneous and exercise-induced GH secretion in postmenopausal women and reduced Sm-C levels. The mechanisms of GH elevation by estrogen may include both central effects and a negative feedback linkage to reduced plasma Sm activity.