Miscibility of Tremolite and Hornblende in Progressive Skagit Metamorphic Suite, North Cascades, Washington     

MISCH  PETER; RICE  JACK M. 《Journal of Petrology》1975,16(1):1-21

Tremolite-hornblende relationships are reported for high gradeand relatively low grade zones within the Barrovian type SkagitSuite, using rocks that range widely enough in composition toqualify as potential hosts of any calcic amphibole from practicallyA1-free to A1-rich types. Out of over 100 samples analysed bymicroprobe, 22 representative analyses are listed, with end-membercalculations. In the sillimanite grade core of the Skagit Suite, calcic amphibolesfrom various, commonly metasomatized metamorphic ultramaficsand genetically related hornblendites, from amphibolites, schistsand gneisses, and from some metamorphosed impure dolomites showcontinuous solid solution between tremolite and highly aluminoushornblendes ranging from almost Fe-free to moderately Fe-richtypes. A1_total/A1^IV is nearly constant and approximates 1?44.The second group of calcic amphiboles studied is from the lower-mediumrange of the epidote amphibolite facies (comprehensively defined),that is, from near and above the oligoclase isograd. Host rocksare variously metasomatized meta-peridotites, and amphibolitesand schists. There is a large compositional gap between analysesof tremolites and of moderately to highly aluminous hornblendes.A1_total/A1^IV approximates 1?73 both in the tremolite and hornblendefields, as against 1?44 at high grade. Fe-poor hornblendes,such as are stable at high grade, were not found at the lowergrade. At both grades, Al shows good overall correlation withNa+K, with A occupancy, and with Ti (with more Ti at high grade). The compositional gap between tremolite and hornblende analysesfrom the lower grade rocks does not necessarily define a solvusbecause no tremolite-hornblende pairs were found. Rather, theanalyses provide outer limits on the possible width of a solvusat this grade (T). However, certain data suggest that a truemiscibility gap not only exists but probably is not a greatdeal narrower than the gap between the actual analyses. Besides,the proposed restriction on A1/Fe ratios at this grade wouldreduce the probability of finding tremolite-hornblende pairs.  相似文献