We report experimental rate coefficients for the energy-pooling collis
ions Cs(6P(1/2))+Cs(6P(1/2))-->Cs(6S(1/2))+Cs(nl(j')) and Cs(6P(3/2))Cs(6P(2/3))-->Cs(6S(1/2))+Cs(nl(J')) where nl(J')=7P(1/2), 7P(3/2) 6D(
3/2), 6D(5/2), 8S(1/2), 4F(5/2), or 4F(7/2). Atoms were excited to eit
her the 6P(1/2) or 6P(3/2) state using a single-mode Ti:sapphire laser
. The excited-atom density and spatial distribution were mapped by mon
itoring the absorption of a counterpropagating single-mode ring dye la
ser beam, tuned to either the 6P(1/2)-->8S(1/2) or 6P(3/2)-->7D(3/2,5/
2) transitions, which could be translated parallel to the pump beam, T
ransmission factors, which describe the average probability that photo
ns emitted within the fluorescence detection region can pns through th
e optically thick vapor without being absorbed, were calculated for al
l relevant transitions. Effective lifetimes of levels populated by ene
rgy-pooling collisions are modified by radiation trapping, and these f
actors were calculated using the Molisch theory. These calculated quan
tities have been combined with the measured excited-atom densities and
fluorescence ratios to yield absolute energy-pooling rate coefficient
s. It was found that the rate for production, in all cases, is greates
t for 6D, but that 1/2-1/2 collisions me significantly more efficient
than 3/2-3/2 collisions for populating 7P. It was also found that 7P(1
/2) is populated two to three times more efficiently than 7P,, in 1/2-
1/2 collisions, but that the 7P fine-structure levels are approximatel
y equally populated in 3/2-3/2 collisions.