A large number of complex and interrelated microphysical processes are involved in the formation of solid and liquid precipitation particles within clouds. These have received detailed attention in recent years and, with certain important exceptions, a moderately accurate quantitative description now exists of the growth and interactions of cloud particles, from their original formation on condensation or sublimation nuclei to their removal from the cloud as large precipitation elements. The objective of this article, which does not include a consideration of the macrophysical, dynamical approach to cloud physics, is to synthesize current knowledge of these microphysical processes and to establish which major problems have been effectively solved and which require appreciable further attention.

The basic physics of the three heterogeneous nucleation processes occurring inside clouds is now well established, although in the case of the formation of the ice phase some more specific information is required. Accurate quantitative predictions of the nucleating activity of artificially produced smokes are gradually emerging. The calculated growth rates of a population of cloud droplets growing by condensation within a rising parcel of air agree well with observation only if mixing between the parcel and its environment is considered. The efficacy of the collision and coalescence mechanism which supplants the condensation process is extremely sensitive to the values of the collision efficiencies for small cloud drop-lets. The most recent calculations suggest that the 18 μm threshold is incorrect and that droplets of smaller radii can collide. This point and the possible importance of electrical forces in triggering the coalescence process require further experimental study. The combination of the stochastic approach with recently determined collision efficiencies have produced realistic computations of the growth of populations of cloud droplets by means of coalescence. The extremely rapid growth rates observed in certain situations are probably a consequence of electrical forces. The majority of conditions under which raindrops become unstable while falling through a cloud have been established and are supported by a reasonable theory. Experimental studies of the movement of growth layers across an ice surface suggest strongly that the habit of ice crystals is controlled by variations in the surface migration distances of molecules, but further work on the prism faces of ice is required before a definitive conclusion can be drawn. Collection efficiencies of ice crystals have been measured and can be significantly increased by strong electric fields. A good quantitative understanding now exists of the heat and mass exchange of hailstones of simplified geometry, permitting accurate predictions to be made of the various growth regimes occuring in natural conditions. The dominant factors governing the density and structure of hailstones have also been established. However, the microphysics of the individual interactions occurring during the riming process is not well understood. None of the explanations that have been presented for the formation and multiplication of ice crystals within clouds at temperatures close to 0 °C have been rigorously explored, although electrofreezing is supported by appreciable evidence and may be responsible for the primary crystals in certain situations.