Abstract
It is known that thermal conductivity of hydrogen gas over 4500 K derived from theories much deviates from conductivity determined through experiments. This paper points out that the conductivity determined by experiments is not directly measured, and that the method is inconsistent. The inconsistency results from the use of Elenbaas-Heller equation despite of strong nonequilibrium of the used arc plasmas. The real thermal conductivity is estimated through examinations of temperature profiles and E-I characteristics derived from the two-temperature theory and those measured by Steinberger. It is reasonable that the real thermal conductivity lies between two curves obtained by theories and by experiments.