The main purpose of this paper is to verify that application of sampling method to the First Japanese Population Census is a study presenting “mathematization of statistics” in Japan. The word, “mathematization of statistics” is my special term for analyzing the historical process of statistics. The three problems addressed in this paper are as follows. First, I analyze the common points among various researches concerning “mathematization of statistics” in the world of around 1920. Second, I clarify details on the sampling theory and practice by Toyojiro Kameda, because the application of sampling to First Japanese Population Census was realized by Kameda. Finally, I try to evaluate his contribution to the development of statistics in Japan from the viewpoint of “mathematization of statistics”.
