On the basis of a a careful analysis of resonance line formation (both
for singlets and doublets) in structured winds, present time dependen
t models of the line driven winds of hot stars (Owocki et al., this vo
lume; Feldmeier, this volume) are shown to be able to explain a number
of observational features with respect to variability and structure:
they are (in principle) able to reproduce the black and broad troughs
(without any artificial ''turbulence velocity'') and the ''blue edge v
ariability'' observed in saturated resonance lines; they might explain
the ''long lived narrow absorption components'' often observed in uns
aturated lines Fit high velocities; they predict a relation between th
e ''edge velocity'' of UV-lines and the radiation temperature of the o
bserved X-ray emission. As a first example of the extent to which theo
retical models can be constrained by comparisons between observations
and profiles calculated by spectrum synthesis from structured winds, w
e show here that models with deep-seated onset of structure formation
(greater than or similar to 1.1R) produce resonance lines which agree
qualitatively with observational findings; in contrast, the here pres
ented models with structure formation only well out in the wind (great
er than or similar to 1.6R) fail in this respect.