RHEOLOGICAL CHANGES IN EMULSION PRODUCTS WHEN AGED 603 been published in the literature over many years. You will find that when you get on to the powers of the volume concentration the multiples involved vary from paper to paper. This, I think, is essentially due to the fact that people have ignored the effect of particle size. It is only in the past few years, in work such as that of Saunders of the Dow Chemical Company in the U.S.A., that the influence of particle size on viscosity has been clearly demonstrated. He worked with extremely dilute systems, but nevertheless showed that if you have very small particles, about 1 /• down to 1,000 A units, then even in the range where Einstein's original equation is supposed to apply you are getting a particle size effect, and this must be taken into account. The only comfort I can give so far is the fact that equation (V), which we developed by analysing our own data and that of other people, does seem to apply to practically all viscosity data that we can find where the particle size is specified fairly accurately. In addition, the question of particle size distribution has also to be considered. You can obtain the same mean particle size with different size distributions, and you can run into trouble because of this. What we have tried to do is to get a rough assessment of the effect due to the sub-microscopic size particles, because in any way that you work out your mean size, apart from using the simple arithmetical mean, you get involved in powers of the particle size. When you are dealing with sub-microscopic size droplets, once you start taking powers of their diameters they reduce to insignificance. For the moment the best that we can do is to get a rough assessment of the visible sub-microscopic size particles by assuming a mean diameter of 0.25 !•, work out the mean size for the larger droplets in the normal way, and then take an arithmetical mean of the two values. By doing this we find that we can allow for the influence of these sub-microscopic size particles to a certain extent. This particle size effect has, in fact, been virtually ignored over the years, with the result that the interpretation applied to published literature is completely incorrect, because many of the effects which are supposedly present are due, in fact, to variation in particle size and nothing else. MR. M. J. THORNTON: The postulated instability of globules smaller than about I).5/• is presumably the main reason for viscosity changes on ageing. What theoretical background is there for this idea and what influence has the amount and type of emulsifying agent on the stability of emulsions ? THE LECTURER: These droplets are responsible for viscosity changes only in the very initial stages of ageing as I mentioned, they disappear fairly quickly, and you get quite sharp viscosity changes during this period. Subsequently the main factor is the coagulation of the larger size droplets. The whole idea has been discussed in a recently published paper (13). A viscous type emulsifying agent will increase the viscosity of the continuous phase and therefore decrease the rate at which the globules will come together. The actual coalescence, as distinct from flocculation which is only the first stage of coagulation, depends very much on whether your emulsifying agent forms a layer which is one or more molecule thick round the droplets. For instance, for a material such as protein it is very well established that you can get a layer which is many molecules thick around the droplets. We have been able to alter the rate of coalescence of droplets in a dilute emulsion, such as an ice cream mix, containing 10 per cent fat- in-water stabilized by milk protein, by increasing the amount of the milk protein, because this layer of protein around the droplets then becomes progressively thicker.

6O4 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Take the example of w/o emulsions stabilized by the nonionic emulsifying agents Span 80 or Span 85. It is fairly certain that these emulsifiers, irrespective of the concentrations used, form a layer which is only one molecule thick around the fat droplets. Therefore you cannot alter the rate of coalescence by adjusting the amount of emulsifier. You may alter other processes such as flocculation, but certainly not the rate of coalescence once the droplets have flocculated. MR. M. J. THORNTON: IS Dm a volume or weight mean particle size ? What effect did six passes through a hand homogenizer have on this figure and how was it determined ? It can presumably refer to more than one size distribution depending on the spread of globule sizes present. This does not seem to have been taken into account in any of the theoretical equations advanced. THE LECTURER: This depends very much on the volume concentration of dispersed phase. In very highly concentrated systems you will find a progressive decrease in particle size and also a narrowing of the size distribution with repeated homo- genization. If you apply this same technique to a very dilute system you get such efficient dispersion after the first passage through a manually operated homogenizer that the subsequent passes will make very little difference. The effect depends very much on the volume concentration of the dispersed phase. MR. R. L. STEPHENS: Do aged emulsions return to their original viscosity if they are rehomogenized ? THE LECTURER: YOU can not re-emulsify these emulsions once they have broken down. I do not think this is due to hydrolysis of the emulsifying agents in this particular type of system you seem to get some sort of solid precipitate formed at the interface following adsorption of emulsifier. Once this has been displaced from the interface you cannot put it back again. This also applies with protein- stabilized emulsions. You get denaturation following adsorption at the interface therefore the protein has changed its condition and cannot be used again. You can only re-emulsify by introducing some more emulsifying agent. I would never recom- mend making an emulsion in this way, because one has to be very careful about the initial stages of dispersion in preparing an emulsion. You must add only a few droplets at a time of the material which is going to be dispersed, and make sure that this disperses really completely before adding any more rather than bulk intro- duction. A good emulsion is never obtained by adding the liquid to be dispersed in bulk. MR. R. L. STEPHENS: IS there in fact a "no-man's-land" of instability of emulsions between gross particle sizes of the order of 0.5/• and colloidal solutions of a miceliar nature ? THE LECTURE•: A lot of change can be going on in the system without any visible change. This is the point I am trying to make. Particularly when sub- microscopic or very small particles are present, the actual size distribution, and the average size, may appear to undergo very little change- whereas in actual fact there has been a substantial decrease in the number of droplets. You can, for instance, start off with a system containing perhaps a million or more droplets per cc of emulsion. Although you get no apparently visible change in the emulsion, that number might have dropped by one or two orders of magnitude. The only instance I know where systems are fairly stable - and references in published literature