MR. J. S. SAWYER : As the author remarks, if equality of K , and Kh exists it is to be attributed to the correlation of humidity and density fluctuations in the turbulent flow. Such a correlation must be affected by the nature ofthe evaporation process at the gr0unde.g. whetherthe evaporating areas are warm or cool relative to areas with less evaporation. Does this imply that the ratio of K , and Kh found in any experiments such as those reported by the author will be rather dependent on the evaporative properties of the underlying vegetation ?DR. H. L. PENMAN : Salutary experience suggests that although the logarithmic profile is true as a long period mean, it is rarely true over short periods. Perhaps unhelpfully, it seems to me that the combined flux ( H + E) is dictated by energy supply, the partition by surface and atmospheric factors, and the gradient of vapour pressure and temperature simply change to cope with the fluxes imposed.Why were the diffusivities for heat and water vapour not compared directly from the Davis measurements instead of comparing the behaviour of E , and H , at two different sites ? MR. J. L. H. SIBBONS (partly communicated) : Dr. Penman has referred to the hypothesis that the available energy is the overriding factor where evaporation is concerned, the profiles and gradients having to adjust themselves to the prevailing flux regime. However, even in the absence of advection or a limiting supply of water the amount of energy available is likely to differ from one type of surface to another. When the boundary is characterized by a very large roughness length the magnitude of K , in the overlying layers is correspondingly high, and the rate at which vapour can be transported away from the surface tends to outstrip that at which water can be converted by the available energy from the liquid to the vapour phase. Adjustment must take the form of a lowering of surface temperature. This acts both to diminish the capacity to remove water vapour by reducing the specific humidity gradient and to augment the energy supply by cutting down the loss of sensible heat from the surface or, should the fall in surface temperature be sufficiently pronounced, causing sensible heat to be received at the boundary, in which case evaporation proceeds at a rate in excess of that maintainable by the net radiation (Sibbons 1958; 1962). The form of the adjustment should be the reverse when the boundary is aerodynamically smooth or of transitional type. Accordingly the potential evaporation and net radiation should be equal at some point on the roughness length scale which remains to be identified and must undergo displacement with changes in thermal stability.The conditions under which the potential evaporation and net radiation will be equal remain a matter for conjecture. Owen and Thomson (1963) have argued on theoretical grounds, and with the support of experimental evidence that some departure from similarity in transfer mechanisms is to be expected above an aerodynamically rough surface. The reason for this ...