It has long been known that the post-core-collapse globular cluster M30 (NG
C 7099) has a bluer-inward color gradient, and recent work suggests that th
e central deficiency of bright red giant stars does not fully account for t
his gradient. This study uses Hubble Space Telescope Wide Field Planetary C
amera 2 images in the F439W and F555W bands, along with ground-based CCD im
ages with a wider field of view for normalization of the noncluster backgro
und contribution, and finds Delta(B - V) similar to 0.3 mag for the overall
cluster starlight over the range 2 " to greater than or similar to 1' in r
adius. The slope of the color profile in this radial range is Delta(B - V)/
Delta log r = +0.20 +/- 0.07 mag dex(-1), where the quoted uncertainty acco
unts for Poisson fluctuations in the small number of bright evolved stars t
hat dominate the cluster light. We explore various algorithms for artificia
lly redistributing the light of bright red giants and horizontal-branch sta
rs uniformly across the cluster. The traditional method of redistribution i
n proportion to the cluster brightness profile is shown to be inaccurate. T
here is no significant residual color gradient in M30 after proper uniform
redistribution of all bright evolved stars; thus, the color gradient in M30
's central region appears to be caused entirely by post-main-sequence stars
. Two classes of plausible dynamical models, Fokker-Planck and multimass Ki
ng models, are combined with theoretical stellar isochrones from Bergbusch
& VandenBerg and from D'Antona and collaborators to quantify the effect of
mass segregation of main-sequence stars. In all cases, mass segregation of
main-sequence stars results in Delta(B - V) similar to -0.06 to +0.02 mag o
ver the range r = 20 "-80 "; this is consistent with M30's residual color g
radient within measurement error. The observed fraction of evolved-star lig
ht in the B and V bands agrees with the corresponding model predictions at
small radii but drops below it for r greater than or similar to 20 ".