In recent years, a lively controversy has existed regarding the sources contributing to variations of the total solar irradiance (TSI) on time scales of the magnetic activity cycle or longer. On one side are those who claim that the TSI variations can be explained entirely as a consequence of surface magnetic features (sunspots, faculae, and magnetic network). On the other side, a number of researchers claim that variations of the solar luminosity emerging from the solar interior, as a consequence of structural readjustments caused by the changing internal (dynamo) magnetic field, are important; and that indeed, they may dominate on those longer time scales. The case for the first viewpoint has been recently made in Eos; the case for the second viewpoint is presented in this article. The case for the surface origin is based on two premisesc(1) Surface activity dominates the TSI variability on hours to months time scales. The passage of an active region on the solar disk can be clearly seen reflected on the measured TSI. (2) The semi-analytical calculations by Spruit <1982, 1994>, which concluded that structural readjustments in the solar interior could not occur on time scales shorter than the thermal time scale for the base of the solar convection zone, approximately 100,000 years.