95 = 10.4°) (McFadden and Reid, 1982).
Paleomagnetic studies were performed on 18 core intervals from 14 sites along Transect EG68. The aim of these studies was to define the magnetic polarity and to judge whether the retrieved volcanic samples originated from in situ basement or from loose and rotated clasts. Figure F6 shows a Zijderveld diagram of demagnetization. After removal of small secondary components related to viscous and/or drill-induced magnetization, the direction of the characteristic magnetic remanence is defined using standard principal component analysis (Kirschvink, 1980). As the azimuth of the drill cores is unknown, only the characteristic inclination is meaningful. Only one interval did not yield a reliable estimate of the characteristic inclination because of scattered demagnetization results (Hole SEG09B). The demagnetization results are listed in Table T4, giving the characteristic inclination, intensity of the natural remanent magnetization, magnetic susceptibility, and Königsberger ratio. Also given is the interpreted polarity together with comments. The polarity for all samples from Transect EG68 is reversed with a mean characteristic inclination of –67.4° (N = 6, k = 51, 95 = 10.4°) (McFadden and Reid, 1982).
In some cases the shipboard study suggested that the retrieved core was loose clasts, either based on the drillers' observations of a sudden increase in drill speed reflecting a transition from hard rock to soft sediment or from visual inspection showing weathering on the bottom of a clast. All these cases are also recognized by highly diverse characteristic inclinations between individual holes at the same site (Holes SEG06A, SEG07A, SEG08A, and SEG19A) (Table T4). For other sites, the in situ nature of the core was questioned during the shipboard inspection, although no conclusive indication was available (Holes SEG04A, SEG04B, SEG11A, SEG11C, SEG14A, and SEG16A). It was expected that by drilling multiple holes in the same unit at such sites, the paleomagnetic results would provide an indication of whether the obtained core was in situ or not. The results show that for some of these multiple hole sites, the retrieved basalts were not in situ. This is suggested by significant differences in the paleomagnetic results between individual holes (Holes SEG04A, SEG04B, SEG11A, and SEG11C) (Table T4) and/or by characteristic inclinations with intermediate directions (Holes SEG14A, SEG16A, and SEG20B) (Table T4).
