In the X-ray band of the electromagnetic spectrum from about 1 keV to over 100 keV a diffuse emission is observed which is dominated by sources at very great distances. This conclusion is based on the high degree of isotropy of the diffuse flux with respect to local sources of emission such as the Galaxy or the local supercluster of galaxies. Only in the microwave band of the spectrum does a similar condition exist.
The 2.7 K microwave background has been attributed to the original explosion associated with the origin of the universe. Since the diffuse cosmic X-ray background (CXB) is not associated with the very earliest history of the universe, to what stage in the early evolution of the universe can it be ascribed? The best attempts to account for the CXB as a superposition of discrete sources of known luminosities such as clusters of galaxies, Seyfert galaxies, N galaxies, BL Lac objects and quasi-stellar objects (QSOs) have failed by nearly a factor of three. This difference may be accounted for by low-luminosity sources with high surface density, and several lines of argument suggest that the source luminosity functions must flatten below the limiting flux of the Einstein surveys (Elvis, Soltan, and Keel 1984; Hamilton and Helfand 1987; Persic et al. 1989). In fact, this flattening of the source luminosity function has now been observed by the ROSAT satellite (Hasinger et al. 1991). The competing hypothesis, that the remaining emission is contributed by a hot intergalactic medium, seems difficult to reconcile with COBE results (Mather et al. 1990), which do not find any distortion of the blackbody spectrum of the microwave background due to scattering of the microwave photons by this hot gas.
Although
this discrepancy may not seem serious when compared to other uncertainties
often encountered in astrophysics, it is currently considered to be
substantially beyond the uncertainties in the precision of our knowledge
concerning the known classes of X-ray emitting objects and how they behave
at large cosmic distances.
Furthermore, the discrepancy may actually be more severe than is
generally
recognized, given the controversy concerning the intensity of this
background radiation.
Most authors have assumed that its intensity at 2 keV is
3.0 photons 
,
as determined by Schwartz (1979) and Marshall et al. (1980).
However, several other measurements have found an intensity nearly 40%
higher (McCammon et al. 1983, Wu et al. 1991, Garmire et al. 1991).
This has led to
recent
suggestions that the slope of the diffuse background steepens below
2-3 keV.
An independently calibrated measurement of the spectral shape and intensity of the diffuse X-ray background spanning the range 0.5 keV, where it is dominated by Galactic emission, to 10 keV, where it is purely extragalactic, is required to resolve these issues. With sufficiently high spectral resolution and sensitivity, such an experiment could also make a very sensitive search for emission or absorption lines in the CXB that might provide clues about its origin (Schwartz 1991). The diffuse background in this energy band is nearly isotropic, so the pointing direction is not critical. With a few square centimeters of area, a modest solid angle, and several months of observing time, a very good spectrum of the diffuse flux can be obtained. CUBIC satisfies these requirements, and will complement the searches for faint sources providing the bulk of this radiation which will be carried out by ROSAT , ASCA , and AXAF .