A random dot pattern that moved within an invisible aperture was used to present two motions contiguously in time. The
motions differed slightly either in speed (Experiments 1 and 3) or in direction (Experiments 2 and 4) and the subject had to
discriminate the sign of the change (e.g. increment or decrement). The same discrimination task was performed when the two
motions were temporally separated by 1 s. In Experiments 1 and 2 discrimination thresholds were measured with motion durations
of 0.125, 0.25, 0.5 and 1.0 s and mean speeds of 2, 4, 8, and 16°:s. In Experiments 3 and 4 thresholds were measured with aperture
widths of 5 and 20 cm. The discrimination of contiguous motions progressively deteriorated with decreasing duration and mean
speed of motion. For the lowest value of duration the Weber fraction for contiguous speeds was more than three times as the
Weber fractions for separate speeds. For the same low value of duration the thresholds for discrimination of direction of
contiguous motions were only about 50% higher than the thresholds for separate motions. The Weber fraction for contiguous
speeds was ca. three times higher with the smaller aperture than with the larger one, provided the ratio ‘aperture width:mean
speed’ (i.e. the lifetime of the moving dots) was less than 0.3 s. Aperture width did not affect the discrimination of direction of
contiguous motions. The discrimination of contiguous motions is discussed together with the known data for detection of changes
in speed and direction. It is suggested that both, detection of changes in speed and discrimination of the sign of speed changes,
may be performed by a common visual mechanism.