Although felsic veins crosscut all rock types, they appear most frequently to be associated with oxide gabbros. Magnetic susceptibility allows this relationship to be assessed quantitatively. Figure F25 is a histogram of magnetic susceptibility intensities for the 196 felsic veins that crosscut whole-round core at the intervals where they occur in the core, out of 203 altogether. Nearly 60% of the veins have values of magnetic susceptibility >800 x 10-6 MU at the precise intervals where they were sampled and thus have susceptibilities clearly higher than background olivine gabbros and troctolites; 26.5% of them have values >2000 x 10-6 MU and thus correspond to seams of oxide gabbro.
The immediate value of magnetic susceptibility at a vein, however, does not indicate the overall susceptibility of the peak region cut by the vein, which might be more appropriate to consider. Accordingly, I divided peak regions into two types: (1) those containing one or more felsic veins and (2) those without felsic veins. Figure F26 shows weighted variations of each of these plotted vs. depth. Between 500 and 1100 mbsf, peak regions with veins have greater magnetic susceptibility than those without veins. Deeper than 1100 mbsf, there are fewer peak regions, these have lower magnetic susceptibility, and there are fewer felsic veins. Throughout the core, there is no correspondence between peak regions with or without veins and fluctuations in background magnetic susceptibility. Most significantly, magnetic susceptibility data show that felsic veins are most abundant amid oxide gabbros with the highest concentrations of oxide minerals. The intimate association of the veins and very oxide-rich rocks supports the interpretation that granitic liquids were the products of extreme high iron differentiation involving the prior precipitation of oxide minerals, the well-known Fenner trend of igneous differentiation (Fenner, 1929, 1931; Bowen and Schairer, 1935). That many of the oxide gabbros have variable, sometimes extensive, degrees of crystal-plastic deformation, whereas most granitic veins that cross them are undeformed, indicates that the molten material that froze in the veins was squeezed or expelled from nearly crystalline host oxide gabbros while they were being deformed. Some veins were later deformed themselves. This is one aspect of the process of differentiation by deformation described by Bowen (1920), or the synkinematic differentiation of Dick et al. (1991).
The fault at 1100 m in Figure F14 is also seen in a difference in vein frequency above and below it (Fig. F27A). The fairly even distribution of veins in the two portions of the core is surprising, given the uneven distribution of oxide gabbros that are the most important host of felsic veins. Vein volume per 10 m of core (Fig. F26B) is a better indication of where felsic material is concentrated, with peaks corresponding to local flattenings of the slope of the curve (its derivative) in Figure F27A. Differences between the crustal blocks above and below 1100 mbsf are still apparent, and the spikes in vein volume correlate in general with the distribution of oxide gabbros with high magnetic susceptibility (Fig. F4).