Results of earlier experimental and theoretical work (reviewed by Tepley [1965a, b] and Wentworth [1965]) suggest that hydromagnetic (hm) emissions are generated by a plasma instability at 4–8 earth radii in the vicinity of the equatorial plane. The emissions then propagate in the ion‐cyclotron (Alfvén) mode along high‐latitude geomagnetic field lines toward the earth. Part of the signal energy penetrates the ionosphere in or near the auroral zones, and the remainder is reflected back along geomagnetic field lines and is successively amplified upon passing through the equatorial interaction region. In this note we suggest that part of the signal energy that penetrates the ionosphere is transformed (possibly through collisional processes) into the isotropic (fast) hm wave mode and then propagates horizontally in an ionospheric waveguide around the Alfvén (or phase) velocity minimum near the F2 peak. The complete propagation path is illustrated schematically in Figure 1. The suggestion of waveguide propagation is based upon the experimental results presented in the accompanying letter [Wentworth et al., 1966], which indicate hm emission propagation times on the order of seconds between widely separated stations in the same hemisphere.

Results are presented of recent measurements of hydromagnetic emissions made at four Pacific Ocean stations widely separated in both latitude and longitude. The signals are almost always observed simultaneously at the four locations. The repetition frequency of the emission fine structure is found to be highly regular and is the same at all stations for any given hm‐emission event. The structural elements occur simultaneously at stations in the same hemisphere, but the elements are found to be alternately spaced (180° out of phase) at stations on opposite sides of the equator. Hence, energy is being received periodically and alternately in the northern and southern hemispheres. The result is applied to the evaluation of three recently proposed models for the generation of hm emissions; two of the three models are immediately rejected. It is also observed that occasionally the fine‐structure repetition frequency exhibits the presence of a harmonic. This effect, referred to as ‘structure doubling,’ is tentatively interpreted as resulting from a superposition of signals from opposite hemispheres.

The paper deals with preliminary results of a frequency-time analysis of geomagnetic micropulsations in the frequency range 0.5-5 cps. The data from which these results derive were recorded near Palo Alto, California, November 18-21, 1960. At times, oscillatory signals of slowly varying frequency were observed continuously for periods of many hours. Signals of more rapidly varying frequency were also observed, but for shorter periods of time, usually 10 to 30 minutes. Oscillations at the same frequencies were found to occur simultaneously on waveform records obtained at California Institute of Technology stations near Ely, Nevada, and near Isabella and Mount Palomar, California. The oscillations are called hydromagnetic (hm) emissions, since preliminary evidence implies that they are generated either above or high in the ionosphere and are propagated downward by hydromagnetic waves. The period preceding and including the interval during which data were obtained was characterized by solar-flare activity that produced a number of geophysical effects and undoubtedly influenced the nature of the oscillations.

Recent experimental results are presented concerning the occurrence, structure, and frequency-latitude dependence of hydromagnetic emissions (regular oscillatory micropulsations in the frequency range 0.4-7 cps). Results are also presented fo.r the occurrence of no.ise bursts (irregular micropulsations in the same frequency range). Evidence is considered suggesting that both hm emissions and noise bursts are generated by related mechanisms. The experimental results are summarized as follows: (1) Occurrence. First, short hm emission bursts sometimes occur about I min after the sudden commencement of magnetic storms. Second, simultaneous occurrence with X-ray bursts and increased riometer absorption has been noted.(2) Structure. Hydromagnetic emissions tend to occur in distinct frequency bands. The bands are usually characterized by a fine structure consisting of a superposition of repetitive wave trains of a few minutes' duration and rapidly increasing frequency. (3) Frequency-latitude dependence. An inverse relationship has been found between the highest observed hm emission frequency and the geomagnetic latitude at which the signal is observed.In this paper we consider micropulsations of still shorter periods, that is, signals in the frequency range near I cps. More specifically, a band pass of 0.4-7 cps is employed. Most micropulsations: of the types of interest to us here occur in this general frequency range. However, the actual lower and upper limits of the pass band, 0.4 and 7 cps respectively, were determined by the design parameters of the detection equipment [Tepley, 1961b].The micropulsations are placed in two general categories. SignMs in one of the categories, referred to below as 'noise bursts,' are closely related to micropulsations of longer periods. Sig-nMs in the other category, which are the primary concern of this paper, are les. s clo.sely related to other types of geomagnetic fiuctuatio.ns. We refer to micropulsations in the latter cate-3317 3318 TEPLEY AND gory as 'hydromagnetic emis'sions,' since it has been demonstrated in separate studies that the properties of the signals are consistent with their generation either above or high in the ionosphere and their subsequent propagation downward through the ionosphere by hydromagnetic waves [Wentworth, 1961; Tepley, 1962] (also private communications from H. Benioff and G. Bodvarsson concerning unpublished work at California Institute of 'Technology). The term 'emission' was introduced to suggest an analogy between these signals and the so-called 'VLF emissions,' which are generated above the ionosphere [Gallet and Helliwell, 1959; Gallet, 1959]. The terms 'hm emission' and 'noise burst' are defined below. In the appendix, they are again considered in relation to nomenclature (pearls, SIP, IPDP, type A and Wpe B oscillations, and solar whistles) currently employed by other workers to describe signals in the same general frequency range.

A review is presented of so me of th e more important prope rti es of Pc 1(0.2-5 cIs) e missions observed at middle and low lati tud es . Special attention is give n to fin e s tru c ture d regular osci ll ati ons referred to b y various workers as pearls, tYIJe A osc ill a tio ns, and h ydromagne ti c (hm) e mi ss ions. Th e fo ll ow· in g aspects of th ese oscillations are d isc ussed.(a) Signal appearance. Th e e mi ss ions are co nsid ere d fro m both th e ir a mplitud e· time (wave· fo rm) appearance observ ed o n c ha rt reco rd s a nd th e ir freq ue ncy-tim e ({t) c ha rac te ri s ti cs observed on so nagra ms. The various t ypes of {t fin e s tru c ture are discussed (ri sing a nd fa lling freq uency e le· me nt s, fan s ha ped e le me nts, e tc.).(b) SimuLtaniet y of ocwrrence at widely separated locations. A hi g h degree of similarity is often found in th e ap peara nce of {t s tru ctura l e leme nt s of hm e mi ss ions reco rd ed simult a neo usly a t wide ly se parate d s tati ons. Attention is give n to the time-s hifts be twee n th ese e le me nts at st ati ons in th e sa me he mi sp he re a nd in oppos it e he mi s phe res.(c) Time of occurrence. Co rre lati ons are co nsid e red be twee n tim es of occurrence of hm e mi ssio ns a nd other geop hys ica l effects s uc h as c ha rged parti c le even ts, magneti c s torm s, a nd variations of the io nos ph eric para me te r F oF2.(d) Latitude effects. Various latitude dependent e mi ss ion c hara c te ri s ti cs a re di sc ussed. These include latitude variations of e mi ss ion fre qu e ncy, fin e s tru c ture repetiti on pe ri od , a nd signa l a mplitude.In addition to th e aspects of th e Pc 1 emission s outlined above, properties of two oth er typ es of e mi ss ions are brie fl y discussed. One of these signals, referred to he re as a "continu ous e miss ion" also lies in th e Pc 1 category. It is often obsuved co ntinuous ly throu g hout th e ni ght and is c haracterized by a slow variation of J.t c haracte ri s ti cs. Th e other signal, whi c h might be pl aced in a P c 1-Pi 1 tran siti on ca tego ry, is observed durin g magneti ca ll y di sturbe d periods. On so nagram s it is c haracte rize d by an irregularly spaced ri sin g freq ue ncy fin e-s tru c ture. When monitore d aurally on ti me· co mpressed magnet ic tap e (speed-up fa cto r of 1000 to 2000), it is c hara cterized by a so und simil ar to bubbles blown unde r wate r.