An attempt is made to clarify a thought experiment introduced by P. T. Landsberg concerning the relativistic heat flow between bodies in relative motion. It is shown that if the problem is analyzed within the covariant thermodynamics developed by R. Balescu, supplemented by the second law of thermodynamics as proposed here, then such heat flow considerations do not fix the transformation of temperature as Landsberg contends. Instead, the transformation of temperature is left as being purely a matter of definition.