Juergen, Dave, Dugald, et. al.				7/23/00
	My remarks will be addressed to each of of you below, more or less
sequentially. Also, in case Mike H. is not on this List, I've forwarded
this e-mail to him to see if any comments can be extracted from him
regarding Dugald's comments. There in Dallas, he may be in a shelter hiding
from the heat!

Juergen:
	From your comments, it appears to me that you have seen the essence
of Dave's comments. However, for the sake of others not so familiar with
metapelites, let me elaborate a little.
	To see the point Dave made in his abstract and which I re-stated in
an earlier e-mail, construct a schematic Thompson '57 AFM projection for
low to medium-low grades for the Al-rich portion of composition space. It
will consist of an assemblge something like Prl + Cld + Chl (using Kretz
abbreviations). Pg will probably be present along with the Ms, but Bt will
be absent. Now consider several isobaric T-increases. A series of reactions
will occur that involve the breakdown of Prl and then assemblages involving
various combinations of an Alsil (which one, depending on P, but probably
Ky most commonly) and phases like Chl, Cld, St, and Grt. Again, Pg will
probably be present along with the Ms, and Bt will still be absent,
although some complications can occur on this later point (See Frank
Spear's book for some of the possibilities and complications for these
Al-super enriched bulk compositions. One can see the possibilities of
approaching the triple point (TP) with an Al-silicate already present as
one of the phases present in such rocks -- actually, at exactly the P of
the TP, Ky will be the Alsil present.  However, probably for kinetic
reasons an actual TP assemblage might never crystallize -- as Juergen
states in his e-mail.
	Actually, depending upon where one places the breakdown of Prl
relative to the TP, one can see the possibility of getting Ky-bearing rocks
at P's considerably lower than the TP. However, the only occurrence I know
of in which Prl goes directly to Ky is in S. Vermont in rocks described
years ago by John Rosenfeld and they are in a relatively high-P setting and
so P was higher than the TP. My only personal experience with these
Al-super rich rocks involves the metabauxites associated with the Baraboo
quartzite in Wisconsin. There, some work done in collaboration with Chuck
Geiger was at least "suggestive" that the Prl had gone to And as T
increased. Others have also described this Prl = And case, but I can't
recall the specific references - it's late and I'm into my second glass of
red wine.
	The existence of Al-super rich metapelites compared to "normal"
pelites was probably first recognized by Arden Albee in his senior thesis
at Harvard, back about (1957?) (A reference to his GSA abstract dealing
with such is given in some of my very early papers in the '60's. Possibly
because Albee was at Harvard at the time that Thompson was formulating his
'57 Graphical Analysis paper, it focused his attention on the fact that the
VAST, VAST majority of metapelites plot below a line (not necessarily a
"tie line", though it may be one in many rocks) connecting Grt and Chl on
an AFM projection. Such rocks develop Bt, then Grt early in their history
typically leading to the super common assemblage Grt+Bt+Chl with Ms as the
only white mica; the Na-phase will be Ab, not Pg. Most commonly (exceptions
can occur),as T increases, these rocks undergo reactions (none involving
Alsil) to bring in St. Then if St is present, AlSil comes into such rocks
most commonly by Dave's reaction (R1) St+Chl=Alsil+Bt or
[(?)St+Grt=Alsil+Bt(?) (I have very little familiarity with this second one
as it would occur in relatively Fe-rich bulk compositions)], or in St-free
rocks by Dave's (R2) which I presume would be over on the very Mg-rich side
of the AFM projection. All of these reactions occur well above the triple
point (Dugald's comments not with standing). Hence, other than by
polymetamorphism, in the (BY FAR!) most common metapelites Alsilicates are
not stable in the PT vicinity of the TP due to bulk compositional
constraints on the assemblages permissible -- as Dave's absract points out.
	Actually, considering the relative rarity of the Al-super-rich bulk
compositions, it's rather surprising how much effort has been expended on
studying them, both in terms of field occurrences and theoretical "stuff".
An exception to this assertion may be in the context of very high-P (>10Kb)
parageneses. I have the impression that some of the very high-Al minerals
like Cld are more common in such rocks. Certainly Pg is much more common in
such high-P rocks. In the context of metapelites, for tectonometamorphic
purposes the "meat and potatoes" of low-P, high-T and typical Barrovian
sequences lies in the details of what happens in rocks having bulk
compositions that plot below the line connecting Grt and Chl on an AFM
projection.

Dave:
	Sorry I failed to cite in my earlier e-mail your abstract which you
had even sent to me by e-mail. My only excuse is that I so thoroughly
agreed with virtually all of it, even when I first received it, that it was
not something that I kept consciously in mind.
	That said, let me add a couple of comments about the reactions in
your abstract in addition to those made above. Reaction R1 is every bit as
important as you imply, indeed, maybe even more important than implied in
your abstract alone. The importance lies in the fact that it almost
certainly lies in virtual coincidence with the reaction Pg=Alsil+Ms+Ab
(which lies just a little above Dave's R3 reaction), the upper stability
limit of of Pg. For various reasons discussed in some of my white mica
papers, this reaction in nature is virtually a duplicate of the
experimental reaction of Chatterjee in that the phases usually contain very
little on those annoying "extra" components that plague so many other
reactions in nature. Moreover, various petrographic and mineral chemistry
arguments present by my colleagues and me over the years strongly suggest
that this reaction and R1 are, for all practical purposes coincidental in
PT space (assuming aH20 is near 1). Hence, if one accepts Chatterjee's
experimental curve, or calculates the position of it, then it should
esssentially be locating the PT position of R1 also. If someone's
independent calculation of the position of R1 doesn't coincide with their
positioning of the Pg breakdown curve, I'd suggest that they'd better do
some re-adjusting because they are at variance with what is very likely,
based on what is observed to be the case in nature. I think that from the
comments in Dave's abstract that he'd probably have them essentially in
coincidence -- and both passing through a pretty big slice of the Sil PT
field.
	Finally, Dave, your reaction R3 would be seen only in incredibly
Na-rich bulk compositions. It was predicted by Thompson and Thompson (1968)
based on thermo etc arguments, but to my knowlegde only once has the
natural assemblage confirming its existence been found -- by Martin Frey.

Dugald:
	First, let me say that I consider your bathograd concept to be
enormously important, but sadly, still not appreciated to the extent it
deserves. I especially liked it right at the start because I had used this
same concept as early as 1970 in an NEIGC guidebook article to
differentiate the P's of the M2 and M3 events I had defined in W. Maine.
Moreover, in your '78 paper you made direct use of the implications of this
work I'd done and published in '74. From my point of view, it's too bad I
didn't have your clever vision to see the much broader, general
implications you so clearly perceived. I'll return to your bathograds in a
moment.
	Regarding the comments in your e-mail received today, I have three
comments: (1) It all seems a bit too bootstrap-like in terms of activity
models etc. etc. (2) I'm sure the St+Qtz equilibria is important to makers
of theoretical petrogenetic grids, but does it occur in many places in
nature, or anywhere? In nature, it's my impression that most commonly St
gets "whacked" by the terminal equilibrium , St+Ms=Alsil+Bt+Grt etc. (3) In
recent years, Mike H has become a big fan of polymetamorphism. I wonder
what he'd say now about the Picuris rocks.
	Returning to your bathograds and the TP, in fact it's your
bathograd concept that really makes the TP important, taking into account
the PT  curves that define it and their intersections with PT curves like
that for Dave's R1. Considering just that R1 curve and its intersection
with the And:Sil curve and Sil:Ky curve, and a series of isobaric T slices
makes a nice classroom illustration that there are several perfectly
legitimate Sil isograds, but each having a quite different PT significance.
Unfortunately, very few published papers make it clear as to what they mean
when they say Sil zone or Sil isograd.

	In conclusion, the TP is extremely important, but the TP assemblage
is usually irrelevant except with regard to raising a red flag about
possible polymetamorphism or some other complexity.
					Cheers,  Charlie

************************************************************
C.V. GUIDOTTI                                Univ. Phone (207) 581 2153
UNIVERSITY OF MAINE                          Univ. Fax (207) 581 2202
DEPARTMENT OF GEOLOGICAL SCIENCES            Home Phone (207) 947 1388
5790 EDWARD T. BRYAND GLOBAL SCIENCES BLDG.  Home Office Phone (207) 947 6086
ORONO, MAINE 04469-5790                      E-mail <[log in to unmask]>
U.S.A.                                       Dept.Home Page
                                             http://www.geology.um.maine.edu.
************************************************************


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%