It is now more than ten years since the first x-ray emitting stellar corona was discovered (Catura, Acton, and Johnson 1975). Since then, rapid progress has been made in cataloging the incidence and range of stellar x-ray emission, most notably with the Einstein Observatory (see, eg., reviews by Stern 1983Stern , 1984;; Rosner, Golub, and Valana 1985, Haisch 1986 ). For strong coronal sources, such as RS CVn systems, non-dispersive, low resolution spectroscopic observations have been made with the Imaging Proportional Counter (Majer et al., 1986 ) and also at higher spectral resolution (dE/E -~ 10 -20%) with the Solid State Spectrometer (SSS) on Einstein. Typical results of such observations are that highly active stellar coronas (L, ,,~ 1031 erg s -1) are not well modeled by isothermal plasma emission, requiring at least two components with different temperatures, and in some cases, different emission measures ( Swank et al. 1981, Mewe et EL1. 1982 ). For active stars in the Hyades cluster observed with the IPC, single-component, distributed emission measure loop models with flare-like maximum loop temperatures (Tma, "~ 1-2 x 107 K ) are equally viable interpretations of the data (Stern, Antiochos, and Harnden 1986 ). Higher resolution EXOSAT transmission grating spectroscopy has indicated the need for a multi-thermal plasma in active stellar coronae ( Schrijver and Mewe 1986). More recently, a reanalysis of the EXOSAT spectra has suggested that twocomponent distributed emission measure loop models may accurately model the coronae of Capella and Sigma Corona Borealis (Mewe et al. 1986).
The most active stellar coronae also produce flares with 100 -1000 times the emission measure and x-ray peak luminosities of the largest solar flares (White, Sanford, and Weiler 1978, Stern, Underwood, and Antiochos 1983, Montmerle et al 1983, White et al. 1986 ; and reviews by Haisch 1983, Stern 1984). Simple models of such flares suggest that, given their l0 -50 x 106 K maximum temperatures, they are scaled up solar flares involving a several orders of magnitude increase in flare plasma volume. The spatial distribution of such flare plasma -i.e. many small flaring loops over a large area, or very few, stellar-sized loops -is quite uncertain. Attempts have been made to determine the flare size scale using the decay phase of soft x-ray stellar flares, which usually begins as a simple exponential. This has been interpreted as a time scale for loop plasma cooling, either via radiation, conduction, or both ( e.g, Moore et al. 1980, Stern, Antiochos, and Underwood 1983, Haisch 1983, White et al.1986 ).