282 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 10 and 25 per cent was evaluated in various emulsion systems containing odorless mineral spirits as the auxiliary solvent. The emulsion creaming time was decreased in all cases, although it was still sufficient for most purposes. The emulsions containing the sodium chloride were considerably less milky than the corresponding emulsions without sodium chloride. The sodium chloride appeared to affect the ease of dispersion of the water rather than the strength of the interfacial film. The creamed aqueous layers appeared visually, at least, to be stable without an observable coalescence of water drops. I/iscosity of the lf ater-in-Oil Emulsions and Phase Changes The viscosity of the water-in-oil emulsions is a function of the concen- tration and type of auxiliary solvents, concentration of propellents, aqueous phase, and emulsifying agent. As would be expected, the viscosity of the emulsions increases as the portion of the dispersed aqueous phase increases. Electrical conductivity measurements indicate the emulsions to be of the water-in-oil type. Likewise, creaming occurs from the bottom as would be expected with a water-in-oil emulsion where the dispersed aqueous phase is less dense than the continuous phase and rises to the top. Essentially no foaming occurs when the bottles are shaken. As the viscosity of the emulsions increases the separation times become greater as a result of the increased difficulty of movement of the dispersed water droplets. As the concentration of the aqueous phase is increased to 60-80 per cent, inversion of the emulsion from water-in-oil to oil-in-water generally occurs. After inversion the viscosity of the emulsions decreases with a corresponding decrease in separation time. Electrical conductivity measurements indicate the emulsions to be of the oil-in-water type and creaming now occurs from the top with the denser dispersed propellent phase settling to the bottom. The oil-in-water emulsions foam considerably when shaken, in contrast to the water-in-oil emulsions. Where the previous water-in-oil emulsions gave a nonfoaming spray, the oil-in-water emulsions produce a stream that foams on contact. These results appear to be similar to those obtained by Sherman (9), who noted that the viscosities of water-in-oil emulsions prepared with dis- tilled water in mineral oil reached a maximum value between 77 and 82 per cent water. At higher concentrations of water, inversion of the emulsion occurred. The concentration at which inversion occurred was a function of the amount of emulsifying agent present. Occasionally it was observed that of two supposedly duplicate samples prepared with 20 per cent propellent and 80 per cent water, one was a very viscous water-in-oil emulsion while the other sample was the less viscous

AEROSOL EMULSION SYSTEMS 283 oil-in-water type. The separation time of the former was generally greater than an hour with creaming occurring from tile bottom while tile separation time of the latter was less than a mh•utc with creaming occurring from the top. This was probably a case of the formation of dual emulsions (emulsions with the same liquids and emulsifying agents but havir•g opposite phase types) (10), where the type of emulsio• that was obtained depended upon how the bottle was shaken. This phenomenon, where the phase type of the emulsion is determined by the type of agitation, was first observed by Ostwald (11) and later confirmed by Cheesman and King (12). The electrical conductivity of the emulsions was found generally to be a reliable means for determining the phase of the propellent/water emulsions. In most of the cases investigated the water-in-oil emulsions were essentially nonconducting as the inversion point was approached and became con- ducting after inversion had occurred. However, with "12"/"11" (30/ 70)/water emulsions, it was observed that the emulsion containing 40 per cent propellent and 60 per cent water exhibited an electrical conductance intermediate between that of the nonconducting emulsions with 60 per cent propellent/40 per cent water and the conducting 20/80 emulsion. It is possible that in cases of this type, application of the electric current caused partial inversion of the water-in-oil emulsion to take place. This effect with water-in-oil systems has been reported by Dixon and Bennet- Clark (13). Particle Size of Emulsion Sprays At the present time, very little information on the particle size of the emulsion sprays is available. Attempts to determine the particle size of the emulsions at the "Freon" Products Laboratory by the standard tech- nique involving the rotating slide and wind tunnel (14) have not been successful as yet. In the cases studied, the water evaporated before the droplets reached the slide. In the water-in-oil propellent/water emulsions, the particle size of the sprays is probably controlled to a considerable extent by the degree of dispersion of the water droplets in the propellent. If this is true, the particle size from a given formulation will depend upon the method and efficiency of emulsification. An additional factor probably should be considered and that is the effect upon the emulsion stability resulting from the passage of the emul- sion through a valve. In the emulsions containing a high proportion of water, the droplets are close together. It appears likely that during the violent agitation occurring in the emulsions during the passage through the valve and immediately following their exit from the outer orifice, considerably more coalescence of the water droplets would occur with the concentrated emulsions than with the dilute emulsions. If this happens,