192 JOURNAl. OF THE SOCIETY OF COSMETIC CHEMISTS oo 50 O w 80 I I I TRITONX-IO0- ARLACEL 8:3 I ! I ,/.H--2 4 ! , ( • = 84%•) ! I I ! ! ß '0- -- -0 ...... ß I D"O ¸ 0 -- I t 85 90 95 VOL.% OIL PHASE Figure 3. Effect of mixer location on q• (H = clearance between the mixer impeller and the bottom of the vessel) ½)-.) H:¾ (B) H=..4 Figure 4. Illustration of the difference in flow pattern created by mixer location

EMULSION PHASE INVERSION 193 Z 76 74 72 70 68 66( i I I I 20 40 60 80 I00 X h, W'T. % TWEEN 80 INITIALLY IN AQ. PHASE Figure 5. Effect of initial surfactant location on q• (mineral oil emulsion stabilized with Tween 80 - Arlacel 80 at HLB 8) By determining the values of 4 for a series of identical emulsions pre- pared with varying initial surfactant distribution, the effects of initial surfac- tant location were examined. Figure 5 presents a typical result obtained with nfineral oil emulsions stabilized with Tween 80 and Arlacel 80 at HLB 8. X/, is the weight per cent of the total hydrophilic surfactant (Tween 80) initially present in the aqueous phase. X/, = 100 means that the entire Twccn 80 was initially' in the aqueous phase and X•, --• 0 corresponds to a run where the entire Twccn 80 (as well as all Arlaccl 80) was placed in the oil phase prior to emulsification. It is clear that the effect of the initial surfactant location was very pronounced in this system. The increase of ½ with X/, means that as more hydrophilic surfactant is