The design and operation of a microscope freezing stage developed for use at magnifications up to 500x are described. It makes possible studies of low-temperature phase changes such as the freezing of a saline water phase, and hence an estimate of the total salt concentration, in fluid inclusions as small as 10 mu (10 (super -6) mg in weight). The crystal or polished mineral plate containing the inclusions is viewed while immersed in a thermostated heat exchange medium (acetone) circulating rapidly in order to minimize thermal gradients. The stage permits easy operation at temperatures down to -35 degrees C., with electrical control to + or -0.05 degrees C., and to much lower temperatures with manual control. With substitution of silicone oil for acetone, the same stage can be used for heating experiments up to + 250 degrees C. Calibration points in the low range indicate the accuracy of freezing temperature determinations on optimum material to be better than + or -0.1 degrees C. The low relief of ice crystals in water solution places considerable importance on sample selection, preparation, and lighting. As a result of almost ubiquitous and drastic supercooling (metastability), -35 degrees C. is inadequate to freeze most inclusions. Holding at -78.5 degrees C. (acetone + solid CO 2 ) for 30 minutes is generally adequate, although a few samples require extended immersion in liquid N at -196 degrees C. to cause freezing of even a part of their inclusions. Such extensive supercooling is not possible with most surface waters owing to the presence of abundant extraneous solid nuclei for the crystallization of ice. It is believed that the exceedingly slow flow rates for most deep-seated waters trapped in inclusions permitted settling of any such nuclei. Most inclusions are nearly opaque when solidly frozen; with gradual warming they become translucent rather suddenly when the amount of liquid formed is adequate to fill in between the minute solid grains of ice and salts. This is called the first melting temperature and, although inexact, does provide some useful information as to the gross composition of these multicomponent systems. More significance can be attached to the freezing temperature, defined as that temperature at which the last crystal, usually ice, melts in the inclusion, under reversible equilibrium conditions. Reversibility is verified by causing renewed growth of the last small crystal with a slight drop in temperature. As a result of supercooling, this test can be used at any temperature up to, but not including, the actual equilibrium freezing temperature. As long as the last crystal phase to melt is ice, the depression of the freezing temperature provides an approximate measure of the amount of salts present in solution. The combination of chemical analytical data on leachates, giving the ratios of the various salts present in inclusions, with freezing data, giving the concentration of salts, makes possible much more complete characterization of the included fluids than would be possible from either type of data alone.