In QCD, the gluon is associated with a color charge 
and the quark
with a charge 
. The larger color charge of the gluon means that
it is more likely to radiate an additional gluon than a quark, leading to
differences in the expected properties of quark- and gluon-induced jets.
For quark and gluon jets produced with the same energy and under the same
conditions, gluon jets are expected to have a larger mean particle
multiplicity than quark jets [59]. The larger multiplicity of
the gluon jet implies that its particle energy spectrum, known as the
fragmentation function, is softer. A related prediction is that the mean
opening angle of particles in a gluon jet is larger than in a quark
jet [60]: thus the gluon jets are broader. Much experimental
effort has been invested in an attempt to observe these predicted
differences (for a recent compilation, see [61] and references
therein). Before LEP 1, there were experimental indications that gluon jets
were indeed broader than quark jets, based on measurements of the mean
transverse momentum of particles in a jet with respect to the jet axis, or
similar variables. However, contradictory results were published concerning
differences between the quark and gluon jet fragmentation functions, while
no evidence was found for a multiplicity difference between the two jet
types. In general, it proved difficult to obtain conclusive results on
quark-gluon jet differences at facilities before LEP 1 either because biases
were introduced by assuming the gluon jets to be the lowest energy jets in
e
three-jet events or else because there was no event-by-event
identification of gluon jets with a resulting lack of sensitivity.
Due to large event statistics and good detector capabilities, the LEP
experiments have been able to settle the experimental question of quark and
gluon jet differences [62,63]. Three aspects of
the LEP 1 studies allow this success. (1) Symmetric events were selected in
which the quark and gluon jets being compared had the same energy and angles
relative to the other jets, allowing a direct, model independent comparison
of the jet properties. (2) The quark jets were tagged, leading to
identification of the gluon jets with better than 90% purity through
anti-tagging. (3) The anti-tagged gluon jet data were combined algebraically
with the quark and gluon jet data from the untagged, symmetric events,
leading to separated quark and gluon jet measurements with essentially
no biases except from the jet definition. In the first LEP 1 studies,
the quark jet samples were the natural ones for Z
decay, given by the
Z
coupling strength to the individual flavors, corresponding to
roughly 20% d, u, s, c and b quarks. In
later studies, b quark jets and uds quark jets were explicitly selected to
compare to gluon jets [64].
These studies resulted in a confirmation of the qualitative differences between quark and gluon jets given above. Selecting 24 GeV jets in a so-called ``Y'' symmetric event topology, it was shown that gluon jets were 60--80% broader than quark jets as measured by the full width at half maximum of the differential energy and multiplicity profiles [65]. The fragmentation function of the gluon jet was observed to be much softer than that of the quark jet. The mean charged particle multiplicity of gluon jets was found to exceed that of quark jets by 20--25%. Besides the Y events, DELPHI [63] studied 30 GeV jets from three-fold symmetric ``Mercedes'' events and obtained similar results. The comparison of the fragmentation function of quark and gluon jets in Y and Mercedes events shows the expected stronger energy dependence for gluon jets. Extensive comparisons of Monte Carlo predictions to the quark and gluon jet data are presented in [65] and [64]. ARIADNE, HERWIG and JETSET were found to be in good agreement with the measurements. The COJETS agreement was somewhat less good.
ALEPH [66] extended these studies by including a measurement of
sub-jet multiplicities [67]. For small values of the sub-jet
resolution scale, 
(defined using the 
jet finder), the ratio
of the gluon to quark jet mean sub-jet multiplicity was found to be similar
to the hadron level value of about 1.2 discussed above. After subtracting one
from the mean sub-jet multiplicities to account for the contributions of the
initiating quarks and gluons, the sub-jet multiplicity ratio of gluon to quark
jets was observed to reach a much larger value of about 2.0 as 
approached the resolution scale 
at which the jets were defined. The
explanation for this is that the mean sub-jet multiplicity of the quark jets
approaches unity slightly before that of the gluon jets as
. ARIADNE, HERWIG, JETSET and
NLLJET were all found to reproduce the measurement.
Beyond these studies based on symmetric events, ALEPH and DELPHI have examined
quark and gluon jet properties in non-symmetric three-jet event
configurations. The DELPHI approach [63] is to identify gluon jets
in
three-jet events using anti-tagging methods as mentioned above. The gluon jet
properties were compared to those of quark jets with similar energies found in
radiative QED q
events. The qualitative differences discussed
above between quark and gluon jets were observed to be present for jet
energies
between 5 and 40 GeV and were well reproduced by JETSET.
ALEPH [68] introduced a new method to study the multiplicity
difference between quark and gluon jets in three-jet events, by examining the
mean charged particle multiplicity of the entire event as a function of the
energies and opening angles of the jets in the event. Assuming each event to
be
composed of a gluon jet and two quark jets, and that every particle in an
event
could be associated with one of these jets, a fit was made to extract a value
for the ratio of the mean charged particle multiplicity values of gluon
to quark
jets, 
. The result for all jet energies and event topologies
was 
. The fit results were found to agree well with
those from the
symmetric Y analyses when they were restricted to that geometric situation.
Thus the basic differences expected between quark and gluon jets --- a larger
mean multiplicity, a softer fragmentation function and a larger angular width
of gluon relative to quark jets --- are now all well established by the LEP 1
experiments. The QCD models are in good overall agreement with the measured
differences. Future effort in this field at LEP 1 will probably include
studies
of differences in the identified particle rates in gluon and quark jets,
differences in particle correlation phenomena and attempts to reduce the
reliance of the analysis method on the jet definition (as the ALEPH
study [68] discussed above attempts to do). Already, L3 has
presented results which indicate an enhanced 
meson production rate in
gluon jets compared to the rates predicted by HERWIG and
JETSET [50]. This suggests that the models for
gluon jets may need
to be modified to allow for an enhanced production of isosinglet
mesons [69].