We argued that for transport due to dislocation pairs the ratio of the
Hall resistivity to its classical value is given by rho(xy)/rho(xy)0
= 1/(1 + x), x = piepsilon0omega0(2)mL(y)/en(e) ln (L(y)/a) where epsi
lon0 is the permeability of the vacuum, L(y) in units of meter is the
transverse dimension of the sample, omega0 is the average pinning freq
uency due to impurities, n(e) is the density of the electrons. Taking
an experimental estimate of the pinning frequency of 1 GHz, x almost-e
qual-to 10(-18). Thus the Hall coefficient is very close to the classi
cal value. The longitudinal resistivity is activated in character but
the Hall coefficient is not. This is observed experimentally. On the o
ther hand, if transport were dominated by other ''isotropic'' defects
such as vacancies, the Hall resistivity would also be of activated cha
racter.