Appendix 1: Analytical methods and errors

Thirty eight K-Ar and eight 40Ar/39Ar high-precision age determinations were made at the US Geological Survey, Menlo Park, on a total of 22 rocks from the entire volcanic field. Duplicate or triplicate determinations were carried out on 14 samples in order to improve analytical precision. All ages were measured on whole-rock samples selected after thin-section examination. Most of the samples meet the usual criteria for whole-rock dating (Mankinen © Dal-rymple 1972), but some contain minor amounts of glass and a few samples are very glassy. The samples selected for dating were crushed to 0.5-lmm (-18 to +35 mesh). For K-Ar dating aliquots weighing c. 25 g were used for the Ar measurements. A 10 g aliquot was ground to -200 mesh and splits of the powder were used for K20 measurements, which were made in duplicate on each of two separate splits of sample powder by flame photometry after lithium metaborate fusion and dissolution (Ingamells 1970). Ar analyses were by isotope-dilution mass spectrometry using a high-purity (>99.9%) 38Ar tracer and techniques and equipment described previously (Dalrymple & Lanphere 1969). All samples for Ar extraction were baked overnight at 280°C. Mass analyses were done on a 22.68 cm radius, multiple-collector mass spectrometer with a nominal 90° sector magnet, using automated data collection (Stacey etal 1981).

Errors given for the calculated K-Ar ages of individual measurements are estimates of the standard deviation of analytical precision. The errors were calculated using formulae derived by Cox & Dalrymple (1967) and Dalrymple & Lanphere (1969).