6. QUASARS AND GALAXIES AT HIGH REDSHIFT
Gravitational lenses can be used as probes of the sources that
they magnify. In this section, we review what has been learned about
quasars and galaxies in this way, and also describe how lensing can
sometimes interfere with our attempts to study distant regions of
the universe.
As discussed in Section 3.5,
gravitational microlensing leads to significant
flux variation only if the angular size of the source is smaller than
the Einstein radius of the lens. This allows interesting constraints
to be set on the sizes of quasar emission regions
(Sanitt 1971).
Based on the
variability observed in Q2237+031, the linear size of the continuum
emission region in this source is estimated to be 
1015 cm -
the first direct estimation of the size of a quasar
(Wambsganss
et al. 1990b,
Witt et al. 1991,
Webster et
al. 1991).
Interestingly, this size
is not consistent with a blackbody spectrum from a standard accretion
disk model of the quasar
(Rauch &
Blandford 1991)
and suggests the presence of a non-thermal component.
Future multi-color observations should lead to additional information
on the structure of this quasar
(Wambsganss &
Paczynski 1991).
Interesting effects are also possible in the broad line emission,
particularly if it arises in compact clouds orbiting the quasar
nucleus
(Nemiroff 1988b,
Schneider &
Wambsganss 1990).
Differential magnification can cause variation in the line profiles between
macroimages and this can, in principle, be used to probe the cloud
kinematics.
Since the strongest microlensing events, the so-called
high-amplification events (HAE), are due to the source crossing a
caustic, either a fold or a cusp, and since the generic structure of
caustics is well-understood, it may be possible in favorable cases to
invert the light curve to obtain the one-dimensional structure of the
source
(Kayser et
al. 1986,
Grieger et
al. 1986,
1988,
Refsdal 1990,
Grieger 1990,
cf also Refsdal
1966b).
If observations are done
simultaneously from Earth and one or more spacecraft in the solar
system, then it will be possible to distinguish between microlensing and
intrinsic source variability and also to determine the source profile
as well as the lens velocity.
In the really spectacular examples of gravitational lensing of
resolved sources, such as the blue luminous arcs and radio rings, some
parts of the source are magnified by quite large factors. When a
successful model is developed, this automatically provides a
reconstruction of the source with particularly high resolution
information in those parts of the source that lie close to a caustic.
The very fact that the arcs are highly magnified allows spectroscopy
to be carried out on 27m galaxies, three magnitudes fainter than
normal. Several arc redshifts have been thus determined. An even
more impressive example is the arc in Abell 2390, where a velocity
variation of 378 km s-1 has been measured along the length of
the arc
(Pello et
al. 1991).
By combining a knowledge of the velocity width
with the central surface brightness, it might be possible to measure
the Hubble constant, or more realistically, the evolution of the
Tully-Fisher relation
(Soucail & Fort
1991).
Occasionally, a star in the source galaxy may come close enough to a
caustic to be brightened for a few hours by a huge factor of up to
107 or more
(Miralda-Escudé 1991b).
Other examples of
gravitational lensing give more modest magnifications but may still be
useful. VLBI imaging of the core-jet region of a lensed radio quasar
(Q0957+561), HST imaging of the fuzz around a lensed
optical quasar or
its line emitting regions, and the imaging of Lyman-
emission
regions (Q2016+112) are examples of this. It has been
suggested that
some examples of superluminal velocities seen in the VLBI jets of
extragalactic radio galaxies may be the result of gravitational
magnification of more modest velocities (e.g.
Chitre &
Narlikar 1979),
and that rapid variability may be associated with microlensing,
if the sources are intrinsically extremely compact
(Gopal-Krishna &
Subramanian 1991,
Subramanian &
Gopal-Krishna 1991).
