Abstract: Estimation has not yet been systematically investigated with primate. In this experiment, the ability of 2 rhesus monkeys(macaca mulatta) in estimating the numerousness of less than 50 was investigated. In all of the six procedures of the experiment. the delaved-match-to-sample was used to test how accurately the animals could recognize the Standard Number(SN) and Comparative Number(CN). The SN and CN were black-filled circles (dots) that were drawn on white cards. Each number from 9-49 with 2 or 4 apart was represented by 20 individually constructed cards. so each number was represented by 40 discriminable patterns. The fixed patterns were avoided and in order to control the cumulative area, three diameters for black dots were used. The results suggested that. there were two factors. the number distances between SN and CN and the magnitude of SN and CN. affected the numerousness estimation of monkeys. The two monkeys completed all performances of estimation for numerousness of less than 50 when the number distances between SN and CN were 4 or more than 4(80% correct criterion). While when the number distance was 2, the two monkeys reached criterion of performances of estimation for 9- 25, but not for 27- 49. The author suggested that the monkeys completed a relative numerousness estimation and it was a process of assignment of a numerical tag.

Key words: rhesus monkey, numerical competence in primate, relative numerousness estimation, standard number and comparative number