The United Kingdom has recently set up a new system of crop variety performance trials to meet European Economic Community directives concerning the marketing of seeds. This has led to a reappraisal of statistical practices in statutory variety testing including trial design, methods of data analysis and allocation of resources. The paper describes some of the problems encountered and reports work done so far in resolving them. This work includes the construction of a catalogue of generalized lattice designs. A computerbased system is used for routine data processing; it incorporates least squares procedures appropriate to incomplete varieties x trials tables. Much remains to be done in designing series of trials and interpreting their results.

SummaryThis paper presents the results of analyses of yield variability in more than 500 trials of varieties of spring oats, spring wheat, spring barley and winter wheat. A table of critical percentage yield differences has been constructed for guidance in planning future series of trials.

No abstract

The degree of precision obtained in the quality testing of grass varieties under two systems of management was examined (I) in a single cut, (2) under a system of proportionate bulking of produce from individual cuts during a season, and (3) after pooling two seasons' data from two centres. Crude protein, crude fibre, soluble carbohydrate, and in vitro digestible dry matter were determined. The components of variance were analysed and in the data from single cuts varietal differences in digestibility of 210 were significant (P--0-05). Using data from two seasons at two centres, a high degree of precision was obtained for all attributes in a grazing management system and varietal differences in digestibility of 1-25% were significant. Variation due to location and season was greater in hay cuts taken 10 days after car emergence. In both systems varieties differed more in soluble carbohydrate content than in the other chemical components.

No abstract