# Polarity definition of thermoelectric effect

After a brief search of the internet the polarity definition for thermoelectric voltage is not found.

Let conductor X cross a temperature gradient where temperature A is higher than temperature B.

Thermoelectric voltage is created in lead X between points A and B. How is the polarity defined, and what is the explanation for the voltage?

Tentative thoughts: Larger particle movements at high temperature point A causes pressure on particles so those particles that can move, have potential for moving to the colder region B. If dominated by electrons with negative charge, point A would become positive and B negative. Is this called positive or negative Seebeck coefficient? Is it reasonable to assume change in charge for A minus B decides the polarity.

Experiment performed:

If I measure a known voltage source with reb + lead on the positive terminal and black negative lead on negative terminal, I get positive voltage.

If I measure conductor X with positive reported thermovoltages, with red instrument lead connected to point A and black instrument lead connected to point B, and with measurement leads of copper which has lower reported thermovoltages than conductor X, I get negative voltage. Repeated with conductor Y with negative reported thermovoltages (absolute values larger than copper), and got positive voltage.

So it seems the polarity convention is not what I described in 'tentative thoughts', and against what seem logical to assume.