22 samples of cherts from the Central Pacific (Leg17, Hole 167) were analyzed for B content and d¹¹B by ion probe. We also analyzed 3 samples of chert breccias from the Campanian Mishash Fm., Israel. In all samples d¹⁸O has been previously measured; for the DSDP samples ⁸⁷Sr/⁸⁶Sr, Ge was also determined. In the breccias B concentration was mapped by a-tracks.

d¹¹B of DSDP cherts varies between Ð5.5 and +8ä, indicating a fractionation of 32 to 45ä with respect to present day seawater B. This fractionation is similar to the observed values in opal-A radiolarians (+2 to +4.5ä, Ishikawa and Nakamura, 1993). Assuming no secular variation in d¹¹B_SW our results can be the reflection of a three-stage precipitation-dissolution process of DSDP cherts: the first leading to Opal-A and involving a large fractionation of B isotopes, the second and third resulting in quartz via opal-CT, involving high solid/liquid ratios and hence a small isotopic fractionation.

Alternatively the d¹¹B-Age curve may be a reflection of a secular change of d¹¹B_SW. In that case the curve would be shifted by some 40-50 m.y towards the present with respect to the curve proposed by Lemarchand et al. (2000). Such a shift was suggested for d¹⁸O in DSDP cherts by Kolodny and Epstein (1976).

The chert breccias show a sharp difference in B concentration between fragments (50-90ppm) and matrix (3-20ppm), confirming earlier a-track mapping and the interpretation that fragments were silicified in contact with seawater, whereas the matrix was transformed to quartz in contact with water of meteoric origin. Indeed d¹¹B in fragments is in a similar range as that the marine DSDP cherts, confirming their marine origin. d¹¹B in the matrix partly overlaps with the values in the fragments but reaches lower values (-33ä), suggesting that whereas the water was primarily meteoric, the salts were in part derived from seawater.