The purpose of this study was to find a latent structure of an acquaintance network in our world, or, in other words, to estimate the shortest communication channel between the "starter person" and the "target person" in Japanese society. A modified chain-letter technique was used for the experiment. A total of 200 subjects were randomly selected to be the starter persons in Fukuoka. They were asked to select one of their acquaintances to be an intermediary who would be more likely to approach to the target person who lived in Osaka. The same procedure was repeated until an acquaintance chain reached the target person. There were two target persons, X and Y. X worked in a famous company (F-condition), while Y worked in an obscure company (O-condition). Other variables among target persons, such as age, sex, education, place of residence, position in a company were controlled. It was found that a total number of fifty-five chains reached the target persons, with a mean number of steps equaling 7.2. However, the mean number in the F-condition was 5.5, while that in the O-condition was 9.2. Our world is unexpectedly small. In the former condition, people tended to use information regarding a target person's occupation and alma mater as tactics to select an intermediary. In the latter condition, information regarding a target person's place of residence and the breadth of acquaintances of an intermediary was often used for selecting a friend. There was a strong trend for male subjects to select male acquaintances and female subjects to select female acquaintances. The same applies to age and occupation as well, indicating that there are basic acquaintance networks in Japan such as same-sex, same-age, same-occupation networks.