The sampled lava flows at 63ºN (Holes 989B and 990A) and 66ºN (Hole 988A) are typical tholeiitic basalts with plagioclase, augite, glass/mesostasis (now replaced), Fe-Ti oxides, and occasional olivine in the groundmass (Duncan, Larsen, Allan, et al., 1996). Phenocrysts, when present, are either plagioclase alone or plagioclase, augite, and olivine. Samples for dating were selected from the massive, least altered, least vesiculated, and best crystallized (coarsest grain size) parts of the volcanic units, avoiding secondary veins when possible. Because of the low to moderate content of K₂O (0.24-0.47 wt%), which mainly resides in the groundmass, and given the fine- to very fine-grained nature of the sampled volcanic units, we have primarily dated whole-rock minicores (0.5-0.7 g). In total, 12 minicores and two plagioclase separates were analyzed. The plagioclase separates were prepared by crushing, sieving (100-500 µm fraction), cleaning in a deionized pure-water concentration using a magnetic Frantz separator, and, finally, by microscopic verification of the purity.

Ages were obtained by ⁴⁰Ar-³⁹Ar incremental heating experiments carried out at Oregon State University, using an AEI MS-10 mass spectrometer for the whole-rock minicores, and a MAP 215-50 mass spectrometer for plagioclase (Duncan, 1991; Duncan and Hargraves, 1990; Sinton and Duncan, 1998). Samples and the FCT-3 biotite monitor were wrapped in Cu foil, stacked in a quartz tube, evacuated, and irradiated with fast neutrons for 6 hr in the core of the TRIGA reactor at Oregon State University. Ages are reported relative to 27.84 ± 0.12 Ma for the FCT-3 biotite, which allows direct comparison of radiometric ages with the time scale of Berggren et al. (1995). The measured argon isotopes (⁴⁰Ar, ³⁹Ar, ³⁸Ar, ³⁷Ar, and ³⁶Ar) were corrected for interfering nuclear reactions on Ca, K, and Cl (McDougall and Harrison, 1988) and for mass fractionation. Apparent ages for each temperature step (50º-200ºC intervals) were calculated assuming an initial atmospheric ⁴⁰Ar/³⁶Ar value of 295.5, and reported uncertainty (1) includes analytical uncertainty and the uncertainty in the age of the monitor. Apparent plateau ages are defined when three or more consecutive step ages agree within analytical uncertainty (1) and contain a minimum of 50% of the ³⁹Ar released between 600º-1400ºC. Apparent plateau ages are the average of step ages weighted by the inverse of their variance (1). The inverse isotope correlation diagram (³⁶Ar/⁴⁰Ar vs. ³⁹Ar/⁴⁰Ar), fitted by least-squares regression, is used to assess the origin of ⁴⁰Ar in a sample (Roddick, 1978). In the ideal case, where the isochron age calculated using plateau steps agrees with the apparent plateau age and the ³⁶Ar/⁴⁰Ar intercept is within 1 uncertainty of the atmospheric ratio (⁴⁰Ar/³⁶Ar = 295.5), the apparent plateau age is considered a true cooling age. The goodness of fit (MSWD, or SUMS/[N-2] where N = number of steps) indicates if the scatter about the regression line is solely analytical (e.g., SUMS/[N-2] <2.5) or implies geological disturbance of the measured argon isotopes (McDougall and Harrison, 1988).

Paleomagnetic data have been obtained for all the drilled flow units. Shipboard paleomagnetic measurements were given in the Leg 163 Initial Reports volume (Duncan, Larsen, Allan, et al., 1996). These results have later been confirmed by shore-based measurements, including analysis of natural remanent magnetization, magnetic susceptibility, and anisotropy of magnetic susceptibility, summarized in a data report (Hooper et al., Chap. 10, this volume). The measured argon ages are calibrated to the magnetostratigraphy of Berggren et al. (1995). We also discuss possible correlations with cryptochrons, which are short magnetic excursions (tiny wiggles) in otherwise reversely magnetized magnetochrons identified in fast-spreading ocean basins (Cande and Kent, 1992, 1995).