The basic features of the so-called ‘cloud-model’ (Grossmann-Doerth and von Uexküll, 1971) are discussed. The model permits us to derive approximate values of certain properties of the elements of the chromospheric fine structure by comparing the observed profiles of strong Fraunhofer lines with a simple theoretical formula involving four free parameters. The method is only applicable if two basic conditions are met: (1) The absorption coefficient varies with wavelength according to a gaussian with a broadening parameter that is independent of height, (2) the variation of the source function Sv with height is sufficiently small so that may be replaced by a constant. The validity of these assumptions, in particular the latter one, is discussed. It is shown that in the case of Hα there are reasons to believe that the assumption is valid provided the features are not very close to the limb (sin θ = 0.9 may constitute the border line). There is also some evidence to suggest that the cloud model will not work for the CaII K-line.