Please note that after these results were reported at Quark Matter 2006 and posted on the preprint server it was found that what is reported here as \textit{ elliptic flow fluctuations}, should rather be taken as an upper limit on the fluctuations. Further analysis has shown that fitting the multiplicity dependence of the q-distribution does not enable one to disentangle non-flow and fluctuations. The data from the q-vector distrubution does not, therefore, exclude the case of zero fluctuations. The remainder of these proceedings we leave as they were originally reported.
Categories: nucl-ex