A high-altitude nuclear explosion, called Starfish, over Johnston Island on July 9, 1962, produced a distinct geomagnetic effect at stations all over the world. This effect consisted of two major parts: initial rapid oscillations followed by slower changes. The rapid oscillations began within several seconds after the shot, and were caused by the propagation of ELF (extremely-low-frequency electromagnetic waves) and hydromagnetic waves. The slower changes reached a maximum a few minutes after the shot and lasted about half an hour. The worldwide pattern of these slow changes indicates that they were due not to a single effect but to a superposition of different effects. A quantitative estimation is made of these slow changes, taking into consideration ionospheric variations, and the results of these estimations are compared with the observed changes.

The statistical study of geomagnetic storms initiated by Moos [1910] has been continued by a number of workers from time to time. Such studies by Ferraro et al. [1951], Sugiura [1953], and Forbush and Vestine [1955] have clearly shown that the magnitude ΔH of storm sudden commencements (ssc's) at Huancayo depends strongly on local time, being markedly greater during the daylight hours. Maeda and Yamamoto [1960] have further shown that this daytime enhancement of ΔH; is also found at other equatorial stations where the dip is less than 20°. On the other hand, it is well known [see, e.g., Jacobs and Obayashi, 1956] that AH (SSG) is on the average larger in the polar and auroral regions and decreases with decreasing latitude down to about 15° in geomagnetic latitude. This article contains supplementary material.