by graphite-furnace atomic-absorption spectrophotometry (AA) (O' Leary and Meier, 1990).Unless otherwise stated for a given analytical procedure, the analysts included one sample duplicate, one analysis method blank, and two reference samples in each job of 40 or less samples. The arithmetic mean and standard deviation of reference materials and duplicate samples were calculated in order to estimate accuracy and precision for each analytical method. A given analytical method was generally considered sufficiently accurate if the absolute value of the laboratory mean minus the best defined reported value was less than or equal to four times the estimated within-laboratory standard deviation. Generally, for the major elements (those commonly occurring in concentrations greater than 1 percent), a relative standard deviation (RSD) less than 1 to 2 percent was considered adequate for precision. For minor elements (those commonly occurring in concentrations of 0.1 to 1.0 percent) a RSD less than 5 percent was considered adequate, and for trace elements (concentrations generally less than 0.1 percent), a RSD of less than 15 percent was considered adequate. The quality assurance manual for the USGS Branch of Geochemistry (Arbogast, 1990) contains estimates of typical performance capabilities for different sample matrices and analyte concentrations, as well as specific data concerning the accuracy and precision of the techniques described above.The analyses for all 53 elements for each data set are given in the digital files in Part B of this report. Elements with very few or no reported values above the lower limit of determinations and some elements that were analyzed by more than one analytical method were deleted from the three tables accompanying this part of the report. Thus, of the original 53 elements determined, 11
