Nature 396, 242–245 (1998)

In this Letter, the X-ray Laue patterns used to give the orientation of our crystal were misinterpreted by 45°. With the magnetic field perpendicular to the RuO2 planes, the nearest-neighbour directions in the square flux-line lattice (FLL) are actually at 45° to the Ru–O–Ru directions in the crystal lattice.

Our values of the superconducting parameters λ and ξ still suggest that pairing occurs primarily on the electrons on the γ-sheet of the Fermi surface under the conditions of the experiment. However, the orientation of the FLL, interpreted within Agterberg's two-component Ginzburg–Landau (TCGL) theory1,2, is no longer consistent with γ-band pairing if we take the values of Fermi surface anisotropy used in refs 1, 2, and obtained by recent band structure calculations3. However, other calculations (T. Oguchi, private communication) give a different sign of Fermi surface anisotropy. TCGL theory predictions may be modified by extra anisotropy of the superconducting energy gap3, by substantial pairing on the α- and/or β-sheets4 or by anisotropy in the electron mass enhancement. Within TCGL theory, the orientation of the FLL that we observe is no longer consistent with the basal plane anisotropy of Hc2 at low temperatures in a pure sample5. Hc2 is observed to be larger with the field in a {110} direction than a {100}. The nature of the pairing in Sr2RuO4 may be more complicated than we suggested previously.