Abstract
When a number of treatments are applied, in successive periods, to a single experimental subject or several experimental subjects, there exists the possibility that a treatment may produce not merely a direct effect, in the period in which it is applied, but a residual effect in the following period. In this paper, a few kinds of serially balanced designs which are suitable for the case when residual effects exist are considered and constructions of them are presented. Furthermore, the statistical analysis of designs based on the sequences are given.
