Examinations in the Meiji Era have been described in the history of education administration or the social history of schooling. However, this paper describes them as a part of social relations outside the school (ex. kinship, community) from the point of view of “social exchange of children”. This enables us to use Lévi=Strauss' and Polanyi's scheme of exchange in dealing with the question of children. It analyses 137 autobiographies of the Nikkei's “Watashi-no Rirekisho” column from both qualitative and quantitative points of view.
Firstly, “Yoshi (adoption)” and “Azuke (entrusting)” can be regarded as formation of connection between families by transferring children. Especially, the importance of the mother-side uncle can be understood in Lévi=Strauss' “restricted exchange” model and Polanyi's “reciprocity” model. Secondly, “Hoko (apprenticeship)” and “Terakoya (private elementary school)” can be considered to have the similar function to connect families through they show elements of Polanyi's “market exchange” or “redistribution” model.
Thirdly, “Gakko (School)”, “Ryo (Dormitory)” and “Jitaku (Parents' home)” have deprived children of the area between families, and include them within the gate to mediate the state and each family in both market exchange and redistribution mode of exchange. As a result, “Gakko”, “Ryo” and “Jitaku” have transformed children from a medium of relations between families into a measure of competition. “Parents' home-School” pattern tends to emphasize the importance of examination more than “Others' home-Non school” pattern. This conclusion suggests a new hypothesis on the mechanism of the upheaval of examination competition since the end of the Meiji Era.