On a more rigorous approach to geothermic problems