A variety of analytical techniques was used in order to obtain an accurate analysis of as many elements as possible. The major elements Si, Ti, Al, Fe, Mn, Mg, Ca, Na, K, and P were determined by inductively-coupled plasma (ICP) techniques, the oxidation state of Fe required a separate titration, and H2O+, H2O-, and CO2 have been combined into loss on ignition (LOI). The LOI were not corrected for the O gain caused by oxidation of ferrous to ferric iron. Trace elements Ba, V, Cr, Ni, Zn, Cu, Pb, and Sr were also determined by ICP analysis. Fourteen rare earth elements (REEs), Sr, Ba, Th, Rb, Nb, Y, Hf, Ta, U, Pb, and Cs were determined by inductively-coupled plasma mass spectrometry (ICP/MS). All of the analyses were contracted to commercial laboratories, except for the ferrous iron titrations, which were done at Iowa State University. ICP analyses were obtained from commercial laboratories (Chemex and Acme) and ICP/MS and 18O/16O analyses were obtained from the GeoAnalytical Laboratory at Washington State University, Charles Knaack was the ICP/MS analyst. Accuracy of the ICP analyses was checked by including USGS standards BCR-1, BHVO-1 and NBS flyash 1633a as unknowns among the samples and repeated runs of the commercial laboratory in-house standards. Accuracy of the ICP/MS analyses was checked by a well characterized BCR-1 clone from the same quarry (BCR-P) and other in-house standards used at the Washington State GeoAnalytical Laboratory. Precision of both the ICP and ICP/MS analyses was estimated at 5% to 10% from duplicate runs of standards and several unknowns, except near the detection limit, which was 0.1X to 0.5X chondrite for trace elements run by ICP/MS. Trace elements run by both ICP and ICP/MS appear to have been more accurately determined by ICP/MS analysis and the ICP/MS data are preferentially reported.
The Rb, Sr, Sm, Nd, Pb, and U concentrations for the three samples were measured by isotope dilution on the same solutions that were analyzed for isotopic compositions. An unsuccessful attempt was made to obtain a date on the Hole 900A rocks by 40Ar/39Ar analysis of a plagioclase separate consisting of relatively clean, but totally recrystallized plagioclase. The Sr, Nd, and Pb isotopes were analyzed at the Free University of Brussels isotope facility on a 7 collector VG sector mass spectrometer. The analytical procedure is entirely comparable to the one used for samples from the Ninetyeast Ridge on ODP Leg 121 (Weis and Frey, 1991). References for blank values and standard compositions (NBS Pb 981, Sr 987 and Nd Merck) can be found in Weis and Frey (1991). Mineral compositions were determined with the Iowa State University electron microprobe using well-analyzed mineral standards obtained from various sources.