Conniffe, Browne & Walshe (1970) proposed a form of experimental design and statistical analysis for grazing experiments. The original paper did not include any example of the analysis on real data. This note presents such an example and makes some comments on the statistical validity of the method of analysis.

In a previous publication (Edwards & Duthie, 1971) the metabolizable energy values for broiler chicks of 11 samples of winter-sown Throws M. S. field beans from the 1968 harvest were reported. The samples had been grown at different locations in England which represented a variety of soil types. The mean ‘classical’ and nitrogen-corrected metabolizable energy values were 2–40 ± 0–09 kcal/g and 2–26 ±0-11 kcal/g respectively, with as much as 27 % difference between the highest and lowest individual values.