This paper reviews the present state of knowledge concerning the relativistic particles of solar origin. The analytic techniques for obtaining information regarding the distribution of particles in the interplanetary space from ground‐based observations are described. The general features of pitch angle anisotropy of high‐energy particles are outlined. The concepts of relativistic particle propagation in the corona and interplanetary space are discussed with special reference to the relevant new ideas that have emerged from studies of energetic particle events. The experimental and theoretical arguments that indicate that particles can be accelerated to high energies by shock waves in the heliomagnetosphere are reviewed. Finally, a brief summary of the well‐established ideas and unsolved problems is presented.

Solar phenomena produce cosmic ray intensity variations over a wide range of time scales. The observed flux is modulated and rendered anisotropic as the particles propagate in the solar wind, and it is occasionally enhanced by the sporadic emission of solar cosmic rays. Both quasi‐static phenomena (the long‐term omnidirectional intensity variation and the steady state diurnal anisotropy) and transient fluctuations (disturbed daily variation and Forbush decrease) as well as the spatial distribution of solar flare particles are represented by a theoretical model that prescribes the role of the several solar‐controlled parameters that characterize the electromagnetic properties of the interplanetary medium. Considerable information concerning the ambient conditions has been obtained with spacecraft. However, the in situ measurements are confined to a limited region near the ecliptic plane. Consequently, in some cases, theoretical predictions based upon them are not in accord with observations of cosmic ray intensity variations. Thus the modulations and anisotropies must be treated in a three‐dimensional framework. It is therefore reasonable to attempt to deduce the properties of the relevant inaccessible regions of the heliosphere from observations of the sun itself. These properties can be determined by relating the various cosmic ray phenomena to changes during a solar cycle and from one cycle to another. To this end, data covering at least two solar cycles are studied to determine the solar cycle dependence of the following effects: long‐term variations in the omnidirectional intensity, Forbush decreases, solar diurnal variations, and solar cosmic ray events (ground level enhancement and polar cap absorption).

This paper reviews the present state of knowledge concerning the diurnal and semidiurnal variations in the galactic cosmic ray intensity. The analytical procedures that are required for extracting from the original data the desired information concerning the characteristics of the anisotropies outside the magnetosphere are described. These include corrections for atmospheric fluctuations, the determination of the amplitudes and phases of the daily variations, and their interpretation in terms of the free space anisotropies that give rise to them. The experimental results concerning the 24-h wave, including its long-term variations with periods of one and two solar cycles, and the characteristics of the total dmrnal anisotropy are then discussed. The semidmrnal amsotropy is considered next. Transient anisotropies which manifest themselves as day-to-day variations, recurrence tendencies, cosmic ray storms, and diurnal variation trains, and which can introduce appreciable changes in the spectral parameters, are also of interest. The evolution of theoretical models developed to account for the diurnal and semidiurnal anisotropies, and fluctuations thereof, are then discussed in the light of the experimental results. Finally, the general features that are now well established, and the nature of the remaining problems, are summarized.

Although the method of superposed epochs (Chree analysis) has been utilized for seven decades, a procedure to determine the statistical significance of the results has not been available heretofore. Consequently, various subjective methods have been utilized in the interpretation of Chree analysis results in several fields. The major problem in the statistical treatment of Chree analysis results arises from the fact that in most studies of natural phenomena, data are neither random nor sequentially independent. In this paper, a statistical procedure which takes this factor into account is developed.