POSSIBLE LARGE SYSTEMATIC-ERROR SOURCE IN ABSOLUTE GRAVIMETRY

We have investigated large systematic errors, exceeding several parts in 10(8) (several tens of microgals), in absolute gravity measurements with the aim of reducing the differences in measured gravity values o btained by different kinds of absolute gravimeter. We find that rotati on of a falling object around a horizontal axis may be induced by smal l differences in the moment of release of the object from its points o f support. This may lead to a large systematic error which seems to be common to all types of absolute gravimeter. We propose a simple model for the rotation of the falling object which may be applied to all ki nds of absolute gravimeter and predict errors of as large as 10(-6) ms (-2) (100 mu Gal) which are confirmed by experiment. To reduce the err or it is necessary to minimize bath the area of the contacting surface and the stress applied to it. It is also necessary to monitor the rot ation of the falling object in order to show that there is no systemat ic error in measured gravity values.