THE CONTINUOUS MIXING OF PARTICULATE SOLIDS 25 with simultaneous mixing and perceptible non-uniformity may be found in the drum. These results, therefore, show that the overall performance of 4'0 • o 2-5- • 2.0- --, i-5- 1.0 0.5 o5 po i-5 2.0 2.5 Reduced time, t R Symbol Drum speed rev rain -• o 20 [3 4O --x-- 50 •*-- 60 • 8O • IOO ß 120 Figure ? Effect of drum speed on residence time distribution (C--curves) the system, including the effect of the outer wall, is such as to give a residence time distribution of the material in the exit stream identical to

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS that of perfect mixing with a lag in response. From the physical point of view it is expected that the residence time distributions at low speeds would obey the normal law. An attempt was made to plot the F(t) Vs t on arithmetic normal probability and log-normal probability paper but 0 5 1.0 1.5 2.0 2.5 3.0 Reduced time, t R Symbol Drum speed rev rain -t --o-- 20 --D- 40 --x-- 50 -e- 60 -•- 8o -•- IOO --,-- 120 Figure 8 Effect of drum speed on residence time distribution (F--functions) straight line relationship could not be obtained. This finding is in contrast to that of Sugimoto et al (28) and the probable reason is that in their experiments they used a horizontal drum with no end wall at the outlet end and the overall flow-pattern is different from the present case. In the present work C- diagrams are asymetrical and show more spread of tracer in the backward direction from its initial position. The present finding does not necessarily mean that the "pseudo-ideal" mixing of other materials will take place at the present operating con- ditions rather it is expected that for the same set of material where the