We investigate the manifestation of stripes in the in-plane resistivity anisotropy in untwinned single crystals of La2−xSrxCuO4 (x = 0.02-0.04) and YBa2Cu3Oy (y = 6.35-7.0). It is found that both systems show strongly temperature-dependent in-plane anisotropy in the lightly hole-doped region and that the anisotropy in YBa2Cu3Oy grows with decreasing y below ∼6.60 despite the decreasing orthorhombicity, which gives most direct evidence that electrons self-organize into a macroscopically anisotropic state. The transport is found to be easier along the direction of the spin stripes already reported, demonstrating that the stripes are intrinsically conducting in cuprates.PACS numbers: 74.25. Fy, 74.25.Dw, 74.20.Mn, 74.72.Dn, 74.72.Bk The mechanism of the high-temperature superconductivity is not settled 15 years after its discovery, mostly because it is unclear how best to describe the stronglycorrelated electrons in the high-T c cuprates. It has recently been discussed [1][2][3][4][5][6][7][8][9][10][11][12][13][14] that the electrons in cuprates self-organize into quasi-one-dimensional stripes, which might bring a paradigm shift in our way of understanding the two-dimensional (2D) electronic system in the cuprates. Though the self-organization is interesting in itself, the conventional wisdom suggests that the stripes are destructive to the superconductivity, because charge ordering would normally lead to an insulating state. However, there are intriguing theoretical proposals [10][11][12] that stripes can instead be responsible for the occurrence of the high-temperature superconductivity if they are conducting and meandering -properties that have never been clearly demonstrated before.In this Letter, we report novel in-plane transport anisotropy in the cuprates, which gives direct evidence for the conducting charge stripes in these materials; the temperature dependence and the magnitude of the anisotropy strongly suggest the stripes to be meandering and forming an electronic liquid crystal [11]. The evidence is shown for two representative materials of the high-temperature superconductors, La 2−x Sr x CuO 4 (LSCO) and YBa 2 Cu 3 O y (YBCO), highlighting the universality of the charge-stripe phenomenon in the cuprates. Most notably, the data for YBCO indicate that the charge stripes govern the transport in samples with T c of as high as 50 K, demonstrating that theories of high-temperature superconductivity should inevitably consider the self-organization of the electrons as an integral part.It is fair to say that the majority of the researchers today believe that the stripes are irrelevant to the superconductivity. This general belief comes not only from the conventional wisdom mentioned above but also from the current experimental situation, which can be summarized as follows: (i) strong evidence for the charge stripe order has been reported [1] only for Nd-doped LSCO, where the superconductivity is strongly suppressed; (ii) in other superconducting cuprates, evidence [2][3][4][5][6][7] is reasonably strong f...

Ando, Lavrov, and Segawa Reply: In the preceding Comment [1], Jánossy, Simon, and Fehér (JSF) offered an alternative picture to explain our recent findings [2] and argued that it is unlikely that the charged stripes exist in hole-doped antiferromagnetic YBa 2 Cu 3 O 61x (YBCO). They proposed that the anomalous behavior of the magnetoresistance (MR) is essentially due to an inplane anisotropy in the resistivity of the bulk. Here we show that it is very unlikely that the charge transport through the bulk is taking place in the antiferromagnetic YBCO, and thus the formation of charged stripes in this system is an inevitable conclusion.The staggered magnetization of the antiferromagnetic (AF) YBCO lies in the ab plane and there should be two equivalent easy axes for the spins. When a small number of holes are doped into the CuO 2 planes that have the Néel order, there are four possible phases for the system: (a) the system is a single domain AF (all the spins lie along only one of the easy axes) and the doped holes are distributed over the system, (b) an AF domain structure is formed (with alternating direction of the easy axis) and the doped holes are distributed in the domains, (c) an AF domain structure is formed and the doped holes are confined in the domain walls (in other words, charged stripes constitute the domain walls), and (d) a macroscopic phase separation takes place into the AF regions and doped nonmagnetic ones.The ESR experiment by JSF demonstrates [1] that there is a magnetic domain structure, which gives evidence against case (a). The equivalence of the a and b axes follows also from the symmetry of the MR [2]. Case (d) can easily be eliminated, because in this case the system must contain an admixture of the superconducting phase (remember that YBCO is either an antiferromagnet or a superconductor); experimentally we have never observed such a macroscopic phase separation. We can also eliminate case (b), because the domain walls in this case have a very large energy cost (of the order of NJ, where N is the number of broken spin bonds along the domain walls and J is the antiferromagnetic interaction energy). Such a large energy cost of the domain walls would cause a transition into a phase without domain walls [case (a)] at low temperature. Therefore, case (c) is the only possible phase in antiferromagnetic YBCO given the AF domain structure is observed down to low T [1]. In case (c), the energy cost of the domain walls is grossly smaller than that in case (b) due to the spinless stripes. The above discussion makes it clear that the charge transport through the bulk is unlikely to be taking place in the antiferromagnetic YBCO.JSF argued that an array of charged stripes cannot respond to the magnetic field because the Coulomb repulsion would make it extremely rigid. This is their central reasoning to reject the stripe scenario. However, the Coulomb repulsion does not necessarily make the stripes rigid; actu-FIG. 1. Magnetization of YBa 2 Cu 3 O 6.3 for Hkab.