Klosse and Ullersma (1973) were the first to attempt an analytical description of the nonisothermal vapour transport across a gas-filled enclosure. They considered the limit for plane ( ) vertical rectangular cavities. Their method was somewhat irrational, however, and was thoroughly criticized by Jhaveri and Rosenberger (1982; §3.3.4). For example, they assumed that the temperature would depend only on , on account of the small value of . This is correct (Cormack, Leal & Imberger 1974), but would equally imply , so that only a trivial mass transfer result would be obtained. Further, like Meyer and Kostin (1975; see comments in §3.2.2), they omitted the inertial terms from the equation of motion but retained the advective component of the vapour flux.