The current models of the solar photosphere consisting of a large number of vertical columns of plasma seem to be in contradiction with the new observational data obtained from the spectrograms taken during the third flight of the Soviet Solar Stratospheric Observatory (SSO) on 1970, July 30. As it was mentioned by , the effective spatial resolution on the spectrograms due to the rough grain of the film is two times smaller than that on direct photographs, being about 0'.'4.
In the model of granulation with radial granules the motion of plasma above the granules was considered to be directed upwards and above the granular interspacedownwards. However previously (Krat, 1961 it was detected that even in the disc center the network of Doppler's shifts of the Fraunhofer lines is somewhat shifted with respect to the picture of granulation. This effect, probably due to a certain inclination of photospheric elements, was observed at Pulkovo with spatial resolution of about 1'.'2-1':5. In a paper by an analogous effect concerning the inclination of the magnetic field lines was suspected.
The first estimates of the strong inclination to the horizontal plane of facular elements were obtained by Stoyanova (1970). Afterwards proposed a generalized model of facules being spiralling magnetic ropes with a considerable screw pitch near to 1000-1500 km. The idea of filamentary structure of the photosphere was recently suggested also by .
We consider the main characteristics of the photosphere model to be consistent with the following picture: The photosphere consists of a system of shallow twisted magnetic arcs divided by the regions of faint or practically negligible magnetic fields. This system of magnetic arcs is statistically stable. This means that the interaction of its elements can be sufficiently strong though the whole system can remain a statistic population nearly unchanged. The arcs of opposite polarity can easily annihilate when approaching one another and the arcs of the same polarity can combine in large scale objects and at the same time the large scale objects can split into smaller ones due to their intrinsic instability (as in sunspots). The existence of magnetic arcs explains the highly inclined motion of plasma, if these arcs are contained inside the solar atmosphere. The relative height of different parts of arcs will not exceed several hundred km. Nevertheless even the existence of low magnetic arcs makes the motion of plasma preferably horizontal if this motion is directed by the magnetic field. However even the simple hydrodynamical wind motion due to the horizontal gradients of pressure propagates nearly in the horizontal direction. Therefore large inclinations of Solar Physics 32 (1973) 307-310.