Experiments on disease in plants often use half-leaves as experimental unit
s. The whole leaves are typically in the rectangular array formed try h lea
f-heights (rows) and p plants (columns), where h < p. We specialize to the
case where h = n and p = 2n and there are 2n treatments. If each treatment
occurs once per column and twice per row,we call such a design a semi-Latin
rectangle. Efficient semi-Latin rectangles of such sizes are constructed f
or 2 <less than or equal to> n less than or equal to 10. Guidance is given
on their use.