It is suggested that the experimental data on the solar neutrino flux as me
asured in the existing solar neutrino detectors (e.g. Homestake, Kamiokande
II and III, Gallex and Sage) vary with the solar activity cycle to a very
high level of statistical significance. We have applied the run test and th
e change point test to the nine sets of solar neutrino flux that have been
generated by the Monte-Carlo simulation with production rate and background
parameters that are typical of those in the actual Homestake experiment. H
omestake solar neutrino flux data show anticorrelation with sunspot numbers
from 1970 to February 1994 at a very high level of statistical significanc
e. However, the Kamiokande solar neutrino flux data show correlation with t
he sunspot number data at a significant level. Again it is shown that out o
f nine Monte-Carlo-simulated data only three indicate a variation within th
e period from 1970 to February 1992, but these three Monte-Carlo-simulated
solar neutrino flux data do not show significant anticorrelation with the s
unspot number data. The solar neutrino flux data from Gallex and Sage show
not only variation within the measurement period, i.e. from January 1990 to
October 1995, but are also correlated with the sunspot numbers. The Kamiok
ande solar neutrino Bur data not only show variation from January 1987 to F
ebruary 1995 but are also correlated with the sunspot number data. The vari
ation of solar neutrino Bur data within the solar activity cycle and antico
rrelation/correlation indicates that the solar activity cycle is due to the
pulsating character of the nuclear energy generation inside the core of th
e sun.