In order to answer some of the questions concerning the "inadequacies" of mirror-images in experiments examining co-operative behaviour during predator inspection, several control experiments were carried out. The average positions (distance from predator's tank) of guppies was influenced by the presence of a preadator, but the effect of previous experience with a predator was manifestated only in the presence of predator. The average position of guppies also depended on the mirror-treatment (changing the length and position) but it was influenced only by the length but not the angle of mirror. The relative independence of the position of the mirror observed in the guppies' behaviour challenges the basic assumption (i.e. closer to the predator by long parallel than short angled mirror) is indeed the result of TFT as interpreted earlier. Instead, the effect of mirror treatment appears superpositioned on the predator inspection as an independent variable. In light of these results we discuss the usefulness of mirror systems in testing predator inspection as TFT-like behaviour.