A new possible mechanism of impurity conduction in semiconductors at low temperatures is proposed. The conductivity is thought of as due to the carrier jumps over the Coulomb potential wall from the occupied impurity centers to the empty ones. The activation energy of conductivity and, in the case of strong carrier-phonon interaction, the conductivity itself is calculated and compared with Fritzsche's experimental data for the so-called "${\epsilon }_2$ anomaly" in $p$- and $n$-type germanium.

• Received 27 February 1961

DOI:https://doi.org/10.1103/PhysRev.123.99