The evidence presented earlier by several authors for the substantial disk eccentricity in dwarf novae during their superoutbursts is shown to result either from errors , or from arbitrary , incorrect assumptions .

( 1 ) The evidence for Z Cha and WZ Sge ( Vogt 1981 ) , based on radial velocities measured from absorption components , was an artifact , resulting from miscalculated beat phases .

( 2 ) The evidence for OY Car ( Krzemiński and Vogt 1985 ) and IY UMa ( Patterson et al .
2000 ) , based on the observed dependence of eclipse parameters on the beat phase , involved an implicit assumption that the observed eclipses are pure disk eclipses , which is not true .
In particular , the observed variations of eclipse parameters are likely due to the contributions from the hot spot and from the superhump source , which depend strongly on the beat phase .

( 3 ) The evidence for OY Car ( Hessman et al 1992 ) and WZ Sge ( Patterson et al .
2002 ) , resulting from the analysis of hot spot eclipses , was based on the assumption that the spot distances are identical with the radius of the disk , which is not always correct .
In particular , in the case of eclipses of `` peculiar '' spots ( involving the stream overflow ) , observed at beat phases away from \phi _ { b } \sim 0.5 , the resulting spot distances are smaller that the radius of the disk .

New determination of disk eccentricity in Z Cha , using Vogt ’ s radial velocities measured from emission components , gives e = 0.05 \pm 0.05 .