Experimental data published by Garner, Nissan, and Wood on secondary stresses in a viscoelastic liquid have been analyzed in terms of the superelasticity theory of secondary stresses as developed by Mooney. The theory was slightly modified to be consistent with the reported viscosity rate of shear behavior of the test liquid. With this modification the theory agreed within experimental error with observed normal stresses at various radii on a stationary circular plate mounted parallel to a rotating plate.One reported qualitative experimental result is in disagreement with the theory, but this particular result is questioned. Other theories of secondary stress are briefly discussed.