We have reanalyzed the Voyager radio occultation data for Titan, exami
ning two alternative approaches to methane condensation. In one approa
ch, methane condensation is facilitated by the presence of nitrogen be
cause nitrogen lowers the condensation level of a methane/nitrogen mix
ture. The resulting enhancement in methane condensation lowers the upp
er limit on surface relative humidity of methane obtained from the Voy
ager occultation data from 0.7 to 0.6. We conclude that in this case t
he surface relative humidity of methane lies between 0.08 and 0.6, wit
h values close to 0.6 indicated. In the other approach, methane is all
owed to become supersaturated and reaches 1.4 times saturation in the
troposphere. In this case, surface humidities up to 100% are allowed b
y the Voyage; occultation data, and thus the upper limit must be set b
y other considerations. We conclude that if supersaturation is include
d, then the surface relative humidity of methane can be any value grea
ter than 0.08-unless a deep ocean is present, in which case the surfac
e relative humidity is limited to less than 0.85. Again, values close
to 0.6 are indicated. Overall, the tropospheric lapse rate on Titan ap
pears to be determined by radiative equilibrium. The lapse rate is eve
rywhere stable against dry convection, but is unstable to moist convec
tion. This finding is consistent with a supersaturated atmosphere in w
hich condensation-and hence moist convection-is inhibited. (C) 1997 Ac
ademic Press.