No abstract

Finding Groups in Data: An Introduction to Cluster Analysis. By L. Kaufman and P. J. Rousseeuw. ISBN 0 471 87876 6. Wiley, New York, 1990. xiv + 342 pp. £39.30.

A lye peeler and a knife peeler were used to peel potatoes of jive quality/size grades in several series of experiments. The weight removed during peeling and trimming, and the trimming rate were measured for different peeling times. Less wastage was incurred with the lye peeler than with the knife peeler especially with large size tubers. The processes were costed and lye peeling was found to be more economical with large potatoes, while lye and knife methods had similar costs with small potatoes. It was more economical to peel potatoes of sub-ware quality (i.e. grading rejects) than those of ware quality, because the extra cost oj'trimming and the extra wastage were more than balanced by the lower cost of the raw material. IntroducfionTHE market for whole pre-peeled potatoes is likely to increase substantially, because of the general trend towards convenience and specialisation, the rising cost of suitable labour, and the shortage of such labour in the catering trade.The most commonly used methods of peeling potatoes on a commercial scale are lye, steam, abrasive and knife peeling. Steam is less suitable than lye for potatoes which are to be used whole and not further processed. This is because a heat ring of cooked flesh on the peeled potato occurs which can be minimised in lye pee1ers.l Steam peeling plant is also more expensive to install and steam-peeled potatoes need more manual trimming than lye-peeled potatoes.2 Abrasive peeling is also not ideally suited for potatoes to be used whole, since a rough surface is left on the peeled potatoes which would lead to quicker enzymic browning than would occur on a knifepeeled potato. The abrasive peeler method is less economical than lyepeeling under most conditions.s It was decided, therefore, to study lye and knife peeling, and to compare their efficiency in terms of wastage and costs, for different grades of potatoes.

The sampling distribution of χ² for agreement with Hardy-Weinberg proportions is investigated for three two-allele co-dominant systems. In two systems (MN and Gc) the empirical distributions do not agree with the theoretical one. With PGM₁, however, the empirical distribution is consistent with the theoretical one.