“…The velocities from 6.8-7.2 km/s to 8.0 km/s are too low for unaltered mantle and may result from serpentinization of the upper mantle (e.g., in fracture zones, Detrick et al, 1993, the COT of nonvolcanic rifted margins, Davy et al, 2016, and slow or ultraslow spreading oceanic basins; e.g., the Labrador Sea; Delescluse et al, 2015;Osler & Louden, 1992 or magma intrusion (e.g., in NW Indian Ocean; Gupta et al, 2010). In our final model (Figure 3b), velocities at depths >3 km below the top basement sit outside the envelopes for Pacific and magmatic oceanic crust newly compiled by Grevemeyer et al (2018;Figure 9a well with the velocity profiles in the models of thin oceanic crust overlying serpentinized mantle in the North Atlantic Ocean (Figure 9b; Davy et al, 2016;Funck et al, 2003;Hopper et al, 2004;Whitmarsh et al, 1996) and the exhumed mantle zone in West Iberia (Figure 9b; Dean et al, 2000;Sallarès et al, 2013) and the Central Tyrrhenian Basin (Figure 9c; Prada et al, 2014). Serpentinization of the upper mantle is facilitated when the crust is thin and fractured to allow seawater to flow in (Minshull et al, 1998;Sauter et al, 2013;Shillington et al, 2006).…”