We have used a multiterminal technique in order to measure the (a,b) plane
excess conductivity Delta sigma in several Bi2Sr2CaCu2O8+x single crystals.
We find that the experimental Delta sigma does not follow a simple power l
aw Delta sigma similar to epsilon (-alpha), with epsilon =ln(T/T-c), and th
at it drops faster than the two-dimensional Aslamazov-Larkin law, alpha = 1
, with increasing temperature. In addition, data for samples with different
doping do not scale on a universal curve. We discuss our data in terms of
microscopic and Ginzburg-Landau theories, where high-momentum fluctuations
are either not excited, or phenomenologically cut off. The experimental Del
ta sigma drops even faster than the prediction of the extended microscopic
theory. However, we can accurately describe all our data up to T approximat
e to1.3 T-c with the GL theory, assuming a sample-dependent cutoff value. W
e relate the cutoff parameter to the doping level of our samples.