Recombination lines in gaseous nebulae frequently yield parent-ion abu
ndances that are several times larger than abundances derived from for
bidden lines. One possible explanation for this discrepancy is the pre
sence of temperature fluctuations. We examine temperature fluctuations
in model nebulae by utilizing Peimbert's t(2) parameter. We have run
large grids of models, varying the stellar temperature and the total h
ydrogen density. We consider two abundance sets: The first uses ''typi
cal'' planetary nebulae abundances, while the second examines the effe
ct of increasing the metals and grains by a factor of 3. We also consi
der both a constant density distribution and one which varies sinusoid
ally with radius. We examine the method of deriving t(2) observational
ly, which uses measured [O III] and Balmer temperatures. We find that
this derived t(2) shows no correlation with the t(2) based on the inte
gral definition. We discuss the reasons for this discrepancy, which in
clude nonvalidity of some of the basic assumptions and theoretical and
observational difficulties with the Balmer temperature. We find that,
in high-metallicity objects especially, noncollisional contributions
to [O III] lambda 24363 can significantly affect the derived temperatu
re. We argue that while temperature fluctuations may result in non-neg
ligible abundance corrections in some objects, they are insufficient t
o resolve the abundance discrepancy.