Abstract
The object of the study reported in this paper was to find the proper size, shape, and number of replications of this size necessary for varietal yield trials with rice. 2. With the area constant, the standard error increased as the size of plot was increased. 3. With area in proportion to size of plots, increasing, the size of plot reduced the standard error of the single replicate, the efficiency of plots is the highest in the smallest plots, but it required several more replications than another. 4. The reduction in the standard error is rapid until about 4 th replication, when 6 th the rate of reduction decrcases markedly. 5. The results above mentioned indicate that plots 5 shaku long and 2 row wide replicated four times would be satisfactory.
