The relationship between the observed magnitude of a star and the air-mass takes the shape of a loop. These loops are caused by the zero-point drift of the photometer in use and by the variations of the atmospheric extinction taking place during the observation (Kviz 1979). From observations where the temperature drift of the photometer and of the photomultiplier were recorded, it is evident that the sense of the loops is correlated with the physical temperature of the photo-multiplier. Measurements have been made with the Swiss telescope and the 7-colour Geneva photometer P 7 using pulse counting with PM tube EMI9798 — described in Burnet (1976) and Burnet, Rufener (1979) — at La Silla, Chile.