The solar atmosphere is heated by a flux of mechanical waves propagating in one or more of the modes : acoustic, Alfv6n and gravitational.
The acoustic theory is compared with observational data and found inadequate. First, the theory is based quantitatively on the B~hm-Vitense convection zone model, and large-scale convective motions (supergranulation) and magnetic fields (unipolar regions) show that convection has another form. On the other hand, when granular motions are invoked the energy flux is too small. Second, atmospheric heating is localized in faculae, and enhanced acoustic flux beneath these regions is no longer explicable. Finally, the short periods of 10-30 s invoked recently appear inexplicable. Objections to the gravitational wave heating process are given briefly.
Previous objections to Alfv6n waves as an energy source followed from the belief that fields were generally uniform and of strength ~ 50 G, now known to be incorrect. Models of Alfv6n wave generation are based on (i) granule eddy motions, (ii) overstable oscillations in subsurface flux tubes and sunspot flux ropes, and (iii) supergranule motions, both horizontal and vertical.
The first provides waves which propagate along thin flux tubes oscillating as taut wires in a compressible fluid; they may explain mottles, fibrils and other small emission features. The second may explain the enormous dissipation in spot groups, including flares. The third has been invoked earlier to explain spicules, and may have effects in the solar wind.