On a relation between the indices of solar activity in the photosphere and the corona