In the present paper various conditions for modeling which have not been sufficiently discussed in the previous paper (Part I) are examined, and the relative position of the condition for similarity, U-M/U-N=(Lm/LN)1/3, proposed in the previous paper, is described. Moreover, some obscure points in the previous paper are made clear.
