A Previous experiment on Spacelab 1 [1, 2] has shown that the ability to discriminate between the mass of objects under the weightless conditions of orbital flight (0 g) is poorer than under earth's gravity (1 g): the differential threshold (DL) rose by a factor of about 1.8. There are at least three possible causes: loss of weight information, incomplete adaptation to an altered force environment and the use of suboptimal techniques for accelerating the objects. Mass information is present only during periods of positive or negative acceleration, when the reactive force from the mass of the object stimulates the pressure receptors in the hand. Under I g the object also exerts a continuous downwards pressure, which must be added to the pressure produced by hand accelerations in a particular direction. Thus, on earth both weight and inertial mass are normally sensed together, while in space only mass can be sensed. Incomplete adaptation to an altered force environment can also impair discrimination [3], which should therefore be poorest on initial exposure to 0 g or reexposure to lg , and then improve. It was not possible to measure any adaptation during Spacelab 1, as most tests were conducted in midmission. In postflight, however, the subjects showed raised weight DLs for 2 to 3 days as predicted. It was also noted that subjects with more vigorous shaking techniques showed better discrimination under 0 g. It was decided to investigate this variable systematically, by instructing subjects to make both short shakes and long shakes. (Analysis of amplitude and duration of hand movements from video recordings confirmed that short shakes normally produce higher accelerations.) The apparatus consisted of a box containing 24 balls and a set of record cards. It is described in greater detail elsewhere [1, 2, 4]. The balls (all 30 mm in diameter) rested in holes, and were secured in place by elastic straps. They varied in mass from 50 to 64 g with several duplicates, and were randomly marked with letters of the alphabet. Four sets of apparatus were used, with different letter codes to reduce the chance of subjects learning particular codes. The box could be opened to present two trays (each containing 12 balls) and a fold-out card desk with 40 record cards retained under a clamp. The cards contained 80 pairs of letters arranged in four columns headed alternately Long and Short. The letters in the lists corresponded to masses giving 20 intervals of 2, 4, 6 and 8 g, with the heavier mass equally often first or second, and no letter combination being repeated [4]. The mean mass for both Long and Short columns was 55.0 g. There were eight different random order lists, each repeated five times. The subject secured the opened box to a worktop. In flight he worked in an upright position in relation to the worktop, while on the ground he sat at a table. He picked out the first two balls on the list with the left hand, then used the right hand to shake each ball in turn and replace it in the left hand, with which he returned both to their holes. He marked the heavier ball on the list, guessing if uncertain. He repeated this for all 80 pairs, alternating between long and short shakes, then put the completed card to the back of the pile. The test took about 20 rain to perform on the ground, and about 25 rain in space. The method of shaking was a two-cycle sideways movement for each ball, moving the bent arm freely from the shoulder out to the right with an intended amplitude of about 20 cm for a short shake and 40 cm for a long shake (the distance being gauged by the apparatus). Subjects were trained to standardize the amplitude and duration of their movements as far as possible. Five right-handed crew members of the DI Spacelab Mission (aged 36-44 years) acted as subjects. There were three male payload specialists, and one male and one female mission specialist. They were