There seem to be at least two viewpoints in anthropological studies which are based on observations and measurements of morphological characters. One is the search for affinities between populations or their phylogenetic lines by utilizing the variations observable in morphological characters. The other is clarifying, through analyses of the variation itself, what adaptive significance or function a morphological character has. Here, from the latter viewpoint, I attempted to elucidate the significance of the existence of a certain morphological character through multivariate statistical analyses. As a subject, I chose shovelling, which is often observable as a hollow in the lingual surface of the maxillary incisor. This character most frequently occurs in Mongoloid populations while it is less frequent in others. The question of why this character is almost exclusive to Mongoloid populations started me doing the present study.

This paper is a modified version of a dissertation submitted to the Faculty of Science, the University of Tokyo, for a Doctor of Science degree.

February 1985