THE INVESTIGATION OF DANDRUFF 625 the presence of unsaturated compounds in sebum lipids might be involved in the causation of dandruff this association might help to explain the limitation of dandruff to areas of the skin where the pilosebaceous apparatus is most fully developed. It is hard to believe that normal sebum has a pathological effect on the epidermis and it seems likely that another factor must be involved for example, this might be the route through which a microbiological influence is exerted. The ability to produce epidermal scales on mouse-skin shows promise that further light could be shed on the nature and treatment of dandruff. Since ethyl oleate must be applied continuously to yield the desired response however, medicated shampoos could not be studied in this manner as the irritant would be washed away. An associated effect of continued application of ethyl oleate was hair loss. This is known to occur with numerous unsaturated compounds, such as chloroprene dimers, allyl esters, squalene and vitamin A (9). It was found that, in large doses applied topically to mouse-skin, vitamin A caused scaling similar to that produced by ethyl oleate, suggesting that its epidermal effects were not specifically related to its "nutritional" functions. On the other hand, in small dosage of less than 100 I.U. per day, Lawrence and Bern (10) reported that vitamin A still increased epidermal thickness. It has been postulated (11) that the effect of vitamin A on the maturation of epidermal cells is that, instead of hard keratin, either soft keratin or mucin is formed, depending on the concentration of vitamin A. This is possibly due to a diversion of sulphur metabolism from cystine synthesis to the manufacture of sulphated polysaccharides (12). In our studies, the effect of vitamin A on a harder type of keratin than is found in hairy skin was demonstrated by using mousetail as the experimental tissue. In cross- section mousetail could be seen to have scales of 'hard' keratinizing epidermis, but no granular layer. Between these scales were areas of 'soft' keratinizing epidermis with a recognizable granular layer and con- taining hair follicles. Jarrett and Spearman (13) showed that topical application of vitamin A to the rodent tail led to granular layer forma- tion and epidermal thickening in the interfollicular region and they utilized this finding in the development of a treatment for psoriasis (where the granular layer is lacking). Used concurrently with the corticosteroid triamcinolone to control epidermal hypertrophy, the topical application of vitamin A led to remission of the lesions. Vitamin A does not suggest itself as a suitable agent for dandruff therapy, as large dosage might result in alopecia, but the screening of anti-dandruff agents by looking for inter- follicular granular layer formation in the mousetail might prove fruitful. Since the effects of vitamin A have been shown to depend partly on the natural tendency of a skin region to form 'hard' or 'soft' keratin, it is not

626 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS easy to predict how far the proposed screening arrangements would have relevance in the case of other topical applications. HISTOLOGICAL ASSESSMENT OF DANDRUFF SEVERITY Our technique for the visual assessment of dandruff is rather lengthy and laborious and it therefore seemed appropriate to investigate the possi- bility of correlating a simple measurement of cell nuclei in scale samples with the severity of the condition. A count of the numbers of nucleated and non-nucleated scales was attempted, though it was realized that this would not necessarily correspond to the overall proportions of nucleated and non-nucleated cells. Samples of dandruff scale were distributed over a number of microscope slides, and stained with haemalum and eosin. For counting, each slide was placed on another slide ruled with equidistant parallel lines and viewed at X50 magnification using an eyepiece graticule. The slides were scanned using the ruled lines as guides. As an arbitrary definition of a "scale", a minimum diameter of approximately 0.25 mm was assigned each scale in this category was counted and classified as nucleated or non-nucleated. The percentage of nucleated scales measured in this way was found to range between about 15 and 90%. No correlation with the visual assess- ment was found one reason for this may have been a distinct bias in the samples, since it is more difficult to obtain a representative sample from a head exhibiting little dandruff than from one with heavy scaling. In view of the obvious difficulties in this approach, it might be preferable to seek a more sensitive chemical estimation, such as the assay of desoxy- ribonucleic acid. DISCUSSION Prior to our investigation, only subjective clinical impressions were avail- able for recording the manifestations of dandruff and any attempt to explore therapeutic measures had to be carried out with indeterminate means of verifying the results. It would perhaps have been naive to hope that our studies would solve all the outstanding questions, but reasonably sound lines along which to pursue the problem do seem to have been established. For example, we have demonstrated that clinical study can be conducted systematically and panel tests undertaken with the use of each subject as his own control during a pre-treatment and incidentally, a post-treatment period if enough subjects are available, separate test and control groups can also be used. It is not yet possible to indicate whether the Pityrosporum ovale has a pathogenic role in relation to dandruff. The abundant supply of fatty matter on the scalp where lesions arise, coupled with the lipophilic character