There are many possible explanations for the differences between these two rocks. It is difficult if not impossible to legitimately extrapolate the whole rock chemistry from the characteristics a domino-sized slice of rock, let alone rock chemistry of a locality or a region. Previous authors have noted that the high-pressure metamorphic rocks in the Franciscan often exhibit marked compositional layering, which causes sample size and frequency to be a major concern in validity and usefulness of results.1 The same authors point out, however, that it is the mineral compositions which act as indicators of the Pressure-Temperature conditions at crystalization, and therefore differences in bulk composition may not preclude valid results in this area of study.
Compositional differences do become important when trying to compare the mineral chemistries of two thin sections. As noted by Cloos2, almost without exception, the observed blueschist-facies rocks from the Franciscan have a mafic bulk composition which suggest a basaltic (or gabbroic) protolith. Occasional samples are so silica-rich as to suggest chert or chert-bearing sediments as the protolith, similarly some are significantly carbonate rich, but definite evidence of sedimentary origins is quite rare2. However, it is impossible to rule out sedimentary origins for the compositional layering, whether it is planar or tightly folded, and both are common. Likewise, relic structural features of basalt or gabbro protoliths such as pillows, vesicles, and/or chilled margins of sheeted dikes, are absent in all but the very lowest-grade rocks.
Whatever the compositional variations may or may not have been in the original rock, there are other possible explanations for the variation in SB7, and the lack of variation in RM-B4a. The glaucophane crystals in RM-B4a are in equilibrium with each other, that is, they probably indicate not only consistent bulk composition but also consistent pressure-temperature conditions (at least over the space of the thin section) during crystallization. Recrystallization may continue over a period of time, or occur in multiple episodes, during which conditions may change within a rock. It is possible that the difference between the two rocks may be explained by assuming that RM-B4a was metamorphosed at a stable temperature and pressure for a long enough period to bring all the glaucophane crystals in the rock to equilibrium, and not metamorphosed since them. SB7, on the other hand, may have withstood changing conditions underwhich glaucophane and other crystals were crystallizing at different times, and different crystals may represent a range of conditions and even show the order in which the rock underwent the changes.
The only shortcoming to this explanation is that if SB7 was crystallizing under gradually changing conditions, one might expect to find zoned crystals, which have a gradually varying composition from the center to the rim of the crystal. Very little zoning was observed in any minerals in SB7. It is clear that more research is required in order to satisfy this question.
In addition, it is possible that more in-depth examination of SB7 and collecting more data to average in may cause the standard deviation to fall. The drastic differences could be attributed to small sample size for this sample. This seems unlikely given the magnitude of the observed variation.
In order to improve this study, it will be necessary to extend the range of data collecting by creating chemical maps of areas of the samples in order to detect any subtle zonation which may be present. Extending the analysis to a number of samples from each area may make it possible to hypothesize about the metamorphic histories of the collecting localities with more authority.