In the former paper I discussed the translation problems of children's picture books from a viewpoint of comparative education. The target books I chose for this purpose were "Totte-series" originally written by Gunilla Wolde in Sweden. These books have not only gained world wide popularity amoung younger children, but are also valued very highy both at home and abroad because of their uniqueness and distinguished educational ideas. The main assumption I based myself upon for the discussion comprises that translating activity or adaptation of imported children's books concerns the prevailing standard of good books and good education seriously in the sense that publishers in each coutry are always eager to perceive, for the sake of their successful enterprise, their own parent-purchasers' real demands on "the good for the children." Careful analysis of adaptively translated versions could therefore uncover these real motives and intention over children, which must have much in common with those informal determinants govering inwardly the actual dynamics of an educational system besides formal regulations. Under these premises Japanese version of Totte-series was critically examined in comparison both with the Swedish original and the English version, with reference to the different educational atmospphere each of these 3 countries has. The former study I mentiond briefly above has left some important questions untouched though it might attain its aims to som successful degree. Most of these questions are clearly related to the proposed premises themselves. What I endeavour in this paper is to give empirical evidence and support to these premises through a series of experiments.