“…This observation implies that the canaliforming fluid behaved like liquid water; namely, a fluid with a low viscosity or a high discharge rate Kargel et al, 1994). Previous studies have argued that the fluid could have been a low-viscosity lava with a low solidus temperature, such as carbonatite Kargel et al, 1994;Treiman, 1994;Baker et al, 1997;Williams-Jones et al, 1998), sulfate lava Komatsu et al, 1992Komatsu et al, , 1993Kargel et al, 1994;Baker et al, 1997;Williams-Jones et al, 1998), and mafic or ultramafic lava Komatsu et al, 1992;Greeley, 1993, 1994), or liquid water (Jones and Pickering, 2003). Reconstruction of the high-resolution morphology of canali is likely to help solve this problem.…”