Current observations suggest that the light from blazars is dominated
by emission from a relativistic jet closely aligned with the line of sight.
The broad-band spectrum of this light is consistent with two separate
components. At long wavelengths, a synchrotron component peaks
(in 
F) in the IR-optical regime for FSRQ
and LBL and in the
UV-X-ray regime for HBL. At short wavelengths, a Compton component peaks
at GeV-TeV energies or above; the shape and variability
of this component are much more poorly constrained because of sparse data.
The strong gamma-ray emission confirms that the observed radiation must
be relativistically beamed or else the pair-production opacity would be
too high for gamma-rays to escape.
The same electrons are likely responsible for the emission in both synchrotron and Compton components. The seed photons scattered to gamma-ray energies may be synchrotron photons for weak-lined blazars (BL Lacs) and UV photons from the BLR in strong-lined blazars. This can be tested by direct observation of the correlated variations in these components.
To date, 3C 279 and PKS 2155-304 are two of the three
best-monitored
blazars (the third is Mrk 421). Available data suggest that the
synchrotron emission comes from an inhomogeneous region. Further data
are needed to determine the physical state of the jet, particularly
well-sampled light curves at the highest energies. The ASCA light curves
of PKS 2155-304 and
Mrk 421, for which the X-ray emission is part of
the synchrotron component, both show the hard X-rays leading the soft
X-rays by ~ 1 hour
(Makino et al. 1996,
Takahashi et al. 1996).
The frequency dependence of the lag goes roughly as -1/2 (this
holds approximately for the EUV and UV as well in PKS 2155-304), as
expected if the delays are related to the synchrotron cooling time.
Finally, blazars exhibit no obvious signatures of accretion disks: they do not have big blue bumps and indeed the simultaneity of variations in the optical and UV continuum emission rules out an origin in a standard accretion disk (Urry et al. 1993). On the other hand, that jets are present provides a theoretical argument for disks, since they may be important for jet formation. At the same time, there may be evidence for winds, possibly of material evaporating from an accretion disk. Strong absorption from highly ionized gas has been seen in the EUV and X-ray spectra of a few BL Lacertae objects (Canizares & Kruper 1984, Madejski et al. 1991, Königl et al. 1995). Interpreting this absorption as a wind implies a very massive, very high velocity outflow of highly ionized matter, either in the jet or close to the central engine.
ACKNOWLEDGEMENTS
This paper includes results from a large, ongoing, collaborative effort to monitor blazars at multiple wavelengths. I thank the many participants in these campaigns for helpful discussions, especially Laura Maraschi, Joe Pesce, Elena Pian, Rita Sambruna, Aldo Treves, and Ann Wehrle. Helpful comments were also provided by an anonymous referee. I am grateful to Joe Pesce, Elena Pian, John Godfrey, and Harry Payne for help in preparing the manuscript. This work was supported in part by NASA grants NAG5-2510, NAG5-2538, and NAG5-3138.