a Search for a New Composition-Dependent Interaction: AN Experimental Test of the 'fifth Force' Hypothesis.

We have performed a torsion balance experiment to search for new composition-dependent interactions. We minimized potential sources of systematic error by configuring the torsion pendulum as a symmetric array of four test objects (of two different materials) that were virtually identical in all respects other than composition. We have explicitly tested for thermal, magnetic, mechanical and gravitational effects that could produce spurious results and are confident that our experiment has not suffered from unidentified sources of systematic error. Furthermore, we continuously rotated the apparatus so that any torques on the composition dipole arising from the horizontal component of a composition-dependent force would be modulated at the rotation rate. The experiment was performed on a hillside to increase its sensitivity to forces of intermediate range, such as the 'fifth force' proposed by Fischbach et al. We have compared the horizontal acceleration differences for two material pairs: Be-Cu and Be-Al. Our null results unambiguously rule out the original 'fifth force' conjecture of a Yukawa coupling to baryon number with a strength 100 times weaker than gravity and a range between 10 and 1000 m. We have also established stringent constraints on a more general coupling to a linear combination of baryon and lepton number.