“…Finally, a word about the pseudoscalar W , the weighted relative potential vorticity. In accordance with the potential vorticity theorem of Ertel (1942), the full (unweighted) potential vorticity based on the absolute (rather than relative) vorticity, namely 2 + ω, and potential density can in realistic limits be used as a Lagrangian tracer, and so can the familiar approximation to potential vorticity used by meteorologists and oceanographers in diagnostics work going back to the late 1940s (see, for example, Truesdell (1954), Gill (1982), Hoskins et al (1985), Pedlosky (1987), Schroeder (1988), Haynes and McIntyre (1990), Monin (1990), Kurgansky (1991), Mueller (1995, and Read et al (2002)). Such tracer-like behaviour is clearly not in general a property of W , but this does not detract from the likely importance of diagnostic studies of W , H and S. What might be expected to emerge from such studies is a better understanding of how in fully developed sloping convection the fields of relative vorticity and density equilibrate in the main body of the fluid, and in particular how they are influenced by frictional effects in boundary layers (see above and H89, pp.…”