Studies in the Yield of Tea. I. The Experimental Errors of Field Experiments with Tea

A uniformity trial of 144 plots of mature tea plucked on 42 occasions is examined with a view to determining an adequate technique for field experiments. 2. Yields are expressed as dry matter; the necessity for, and accuracy of, this procedure is examined. 3. Using the latin square and statistical analysis of variance, a maximum diminution of error up to the plot sizes of acre is found; the elimination of positional variance is shown to be inferior to that generally obtained on annual crops in temperate climates, but this inferiority corresponds with similar experience on other tropical crops.