You mention indicators of low pressure, such as cordierite-orthopyroxene. However, Crd-Opx is not restricted to very low pressures and hence would not be critical to define a high-grade hornfels facies. When I say very low P, I do not mean any P below the kyanite stability field, but would distinguish in terms of facies series between what Blatt & Tracy in their textbook call "low-pressure series" and "hornfels series".
Coming back to my earlier question on pressure: the various hornfels facies up to sanidinite facies are separated from the higher-pressure facies by a continuous line or band, for example in the textbooks of Yardley and Blatt & Tracy, which shows a maximum pressure of around 2-3 kbar for these facies. The meaning of this line/band is still not clear to me, after all this discussion. (Sorry for being persistent). By crossing from one facies into another, something major should happen somewhere in an assemblage of common rock compositions. What is it? To be more precise, I particularly question the justification for albite-epidote hornfels and hornblende hornfels facies. (Perhaps there is more justification for pyroxene hornfels and sanidinite facies. I'll leave it to others to burn the bridges here; I haven't seen much of that kind of rocks in the field.)
As far as Ab-Epi and Hbl hornfels facies are concerned, rocks of basaltic composition (on which the facies terminology is based) are not particularly suitable for distinguishing between hornfels facies and their higher P neighbours. Garnet is not a good indicator as many amphibolites that come, for example, from the amphibolite facies proper are lacking garnet completely. Looking at other rock compositions, Kretz, for example, states in his textbook: "The hornblende-hornfels facies is distinguished from the amphibolite facies by the appearance of andalusite in place of kyanite and by the crystallization of abundant cordierite, with only sporadic garnet." However, if the rock compositions trend towards the magnesian side, cordierite + andalusite are quite happy at 4 kbar, taking into consideration minor non-hydrous fluid components and/or some ferric iron in andalusite.
Miyashiro (1994) who discusses the facies issue at length, states "Certain zones of regional metamorphism show virtually the same relationship between the chemical and mineralogical compositions as certain zones of contact metamorphism. Turner (1968, 1981) proposed a series of new facies names for contact metamorphism, distinct from those of regional metamorphism in spite of their close similarity in the relationship between chemical and mineralogical compositions. This is a clear violation of the original concept of metamorphic facies." See also Dugald's last mail.
It must seem rather odd to a student learning the basics of petrology that the hornfels facies terms use minerals that are not critical for these facies. Albite and epidote are abundant minerals in greenschist facies metabasites, hornblende in amphibolite-facies metabasites and pyroxene in granulites.
School of Geological & Computer Sciences
University of Natal