1. INTRODUCTION

An isotropic, apparently diffuse flux of rays was discovered with SAS-2 in the 40 - 200 MeV range [45]. EGRET, improving and extending the SAS-2 result, measured isotropic -ray emission in the 30 MeV - 100 GeV range [38] with F intensity at 1 GeV at the level of 1 keV/(cm²-s-sr), and with F spectral index -0.10 ± 0.03 (Fig. 1a). The diffuse isotropic -ray background consists of an extragalactic -ray background and an uncertain contribution of quasi-isotropic Galactic rays produced, for example, by Compton-scattered radiations from cosmic-ray electrons. The model-dependent Galactic contribution [38, 42, 43, 44], and the addition at some level of heliospheric flux [28, 34], means that the actual contribution from extragalactic sources is somewhat uncertain. The data in Fig. 1a compares the extragalactic diffuse -ray intensity from EGRET analysis [38] with results using the GALPROP model [44], the latter of which requires an extended ( 4 - 10 kpc) nonthermal electron halo to fit the hard ( -0.4) diffuse Galactic -ray emission. For our purposes, we consider the apparently diffuse extragalactic -ray background (EGRB) of [44] as the conservative upper limit for the superposed intensity of any class of -ray sources, with the [38] intensity as an absolute upper limit to the combined residual intensity from all source classes.

Figure 1. (a) Diffuse extragalactic -ray background from analyses of EGRET data, shown by filled [38] and open [42] data points, compared to model calculations of the contributions to the EGRB for FSRQs and BL Lac objects, and total AGNs [14], star-forming galaxies [35], starburst galaxies [46], structure shocks in clusters of galaxies [21, 6], and GRBs [12]. (b) Fitted EGRET and predicted redshift distributions of FSRQs and BL Lac objects [12]. (c) Fitted EGRET size distribution, and predictions for different flux levels [12].

The GALPROP fits [42] to the OSSE-COMPTEL-EGRET Milky Way intensity spectra in different directions toward the Galaxy implies the total -ray luminosity of the Milky Way galaxy. Scaled to 10³⁹ L₃₉ ergs s^-1, the GALPROP analysis gives L₃₉ = (0.71 - 0.92) for the > 100 MeV -ray luminosity of the Milky Way, a factor 3 greater than the value L₃₉ = (0.16 - 0.32) inferred from COS-B observations [7]. Most of this emission is from secondaries created in cosmic-ray nuclear production processes. The Galactic -ray power provides an important yardstick to assess the total contribution of to the unresolved -ray background of cosmic-ray emissions from star-forming galaxies, as described in more detail below.

Every -ray source class makes a different contribution to the -ray background, including transient events below detector threshold, variously oriented relativistic jet sources, and large numbers of individually weak sources. The basic formalism for making such calculations for beamed and unbeamed sources was given in my Barcelona talk [14]. Here I review the various source classes that likely dominate the composition of the diffuse background: blazars and radio / galaxies; star-forming galaxies of various types; rays from structure-formation shocks; and GRBs.