Ohm's law revision

The standard ohmic measurements by means of two extra leads contain an additional thermal correction to resistance. The current results in heating(cooling) at first(second) sample contact due to Peltier effect. The contacts temperatures are different. The measured voltage is the sum of the ohmic voltage swing and Peltier effect induced thermopower which is linear on current. As a result, the thermal correction to resistance measured exists at $I\to 0$. The correction should be in comparison with ohmic resistance. Above some critical frequency dependent on thermal inertial effects the thermal correction disappears.