This study focused on (1) the relation or transition between the measurement and partitive settings for division, and (2) the study of these two interpretations in general case where the remainder is not necessarily zero. Ninety-four second-grade pupils from four classrooms (two in each of two schools) were presented, through individual interviews, with two types of division situations. Each problem of either type was presented in an open or a partially structured setting; four problems, one of each of the possible combinations of type and setting, were presented in eight different sequences. A sequence effect was found, suggesting that, without instruction, working one type of problem interfered with rather than facilitated working with the other type. In particular, partitive problems in open settings were solved with less success if they followed other problems. Questions related to the findings are raised, as they pertain to the sequencing of content, related materials, and activities involved in conceptualization of whole-number division. (MS)