We examine the solar neutrino flux, as measured by the Homestake neutrino d
etector, to search for evidence of a dependence upon the solar latitude of
the Earth-Sun line that varies from 7.degrees 25 south in mid-March to 7.de
grees 25 north in mid-September. Although the flux does not obviously show
any dependence on latitude, we do find evidence for a dependence of the var
iance of the flux upon latitude. When data from 108 runs of the Homestake e
xperiment are divided into four quartiles, sorted according to latitude, we
find that the northernmost quartile exhibits a larger variance than the ot
her three. By applying the shuffle test, we estimate the probability that t
his could have occurred by chance to be in the range 1%-2%.
For more detailed information, we examine a "reconstructed flux" formed fro
m our recent maximum likelihood spectrum analysis. This procedure indicates
that the variance is largest at about 6.degrees 5 north. We also find that
the spectrum of the variance of the reconstructed flux has a notable peak
at 1 cycle y(-1) tending to confirm a latitude dependence of the variance.
We also examine the 12.88 cycle yr periodicity described in our recent pape
r: and find that the amplitude of the periodicity is greater for the northe
rnmost quartile than for the other quartiles. We suggest that these effects
may be attributed to resonant spin-favor precession of left-hand-helicity
electron neutrinos in the magnetic field of the solar radiative zone.