To present one example of the variations among different laboratories in a different way, the FeO and MgO values obtained on the three standard samples are plotted in Figure 1 as well as the calculated mean. The homogeneity of the samples can be tested by comparing the analyses for the portions in Divisions A and B of Sample IC I. As can be seen from Table 4, the agreement between IC IA and IC I B is good, with variations that lie within the SD for either. In further discussions the portions in Divisions A and B are therefore considered identical within the uncertainty of analysis. The results from the different laboratories are in fairly good agreement. However, the standard deviations are larger than in earlier DSDP interlaboratory comparisons (Wright, 1977; Kirkpatrick et al., 1979; Natland et al., 1979 and Staudigel, 1980). There are also systematic deviations from the mean for some of the oxides analyzed by the laboratories (Table 7). The deviations from the mean do not represent deviations from "correct" values, since the calculated mean is based on only 8 analyses. However, they make it clear that care must be taken when data from one laboratory are compared with data from another. Furthermore, they indicate that these data, which were produced as routine analyses, must be considered to have a larger uncertainty than previous data on standard shipboard samples. Differences between different laboratories of 1.5% in SiO2, 1% in Al2O3, 1% in FeO, 0.6% in CaO, 0.5% in MgO and 0.5% in Na2O are larger than satisfactory, in our opinion, even for routine analyses.

Supplement to: Hubberten, Hans-Wolfgang; Splain, Joan H (1983): Interlaboratory comparison of Leg 69 and 70 basalt standards. In: Cann, JR; Langseth, MG; Honnorez, J; Von Herzen, RP; White, SM; et al. (eds.), Initial Reports of the Deep Sea Drilling Project (U.S. Govt. Printing Office), 69, 861-864