I n my previous paper, I dealt with the neutral and nearly neutral genes in a geographically
structured population and obtained approximation formulae for the distribution of gene fre-
quencies in the entire population and in a local population (Maruyama, 1972). I n this paper I
will discuss the fate of the deleterious genes which arise in geographicallystructured populations.
As in the previous paper, I will use two models of population structure : the stepping-stone model
and the continuous model. I n Q 2 of this paper the assumption is made that a colony of a one-
dimensional population exchanges its members with each of its two neighbouring colonies at the
rate ijm and a colony of a two-dimensional population exchanges its members with each of its
four neighbouring colonies at the rate &m,and that the migration in a continuous population is
isotropical and the variance of the migration distance is &'. These assumptions are not essential
but made for simplicity. For further details of the models, see Maruyama (1970, 1971).