Abstract
The ascent velocity of long-lived plasmoids generated under atmospheric conditions to simulate ball lightning was estimated in [Fussmann et al., Phys. Unserer Zeit 39, 246 (2008) and Jegorov et al., Tech. Phys. 53, 688 (2008): Refs. 1 and 2 in the text, respectively], using a rigid sphere model with poor agreement with the experiment. The plasmoids were, however, deformed. Much better agreement is obtained using the Davies and Taylor formula, which describes the ascent velocity of large spherical-cap bubbles.
