These most interesting letters suggest
an idea I should like to offer: the
language we are using is affecting the
way we think. In particular, our
perception of a magma containing a
'liquid' and a 'gas' may be placing
severe, possibly artificial limits on
the phases composing rocks & magmas and
how they interact. Is it possible (being
ignorant of all but the ancient
literature) there may be alternatives to
the two destinations of H2O mentioned in
these interesting letters: a liquid or a
gas.

Fluids

The words 'fluid phase' are very
appropriate in discussing migmatites,
for I believe a migma becomes a magma
when its framework of crystals loses
sufficient structural integrity for it
to flow as a unit, not around or between
units.

Gas

However, the use of 'fluid' and 'gas' in
the context of thermodynamics confuses
me. It is important to distinguish a gas
from reaction products labelel 'H2O' or
'CO2'. As with liquids, one can and may
have many 'volatile fluids' in
thermodynamic equilibrium; but one can
have only one gas. (This distinction
has, admittedly, an artificial feel.)

Volatile fluid

The temperatures & pressures
petrologists work with make it difficult
to distinguish a liquid from a vapor.
Bowen, that tireless supporter of the
anatectic origin of granites, believed
that the fate of a granitic magma,
undergoing eutectic reactions, was
probably a 'hydrothermal solution' rich
(in addition to H2O) in such volatiles
as Li2O, CO2, S, Cl, & F. Even
'hydrothermal solution', he believed,
probably fails to adequately describe
the compositions of that phase
containing most of the H2O released by a
natural system of reactions.

H2O

Held at 700oC & 4000 bars for a month,
95% of a liquid (made in 1958 from
melting the Westerly granite and
saturating it with H2O) evaporated. The
residue precipitated corroded pyroxene &
euhedral garnets. Though the
concentration of H2O in the 'vapor'
could not be measured, Tuttle & Bowen
did not consider it great. This vapor
was dense.

Vapors

The composition of one phase at an
invariant point can be measured from the
other compositions and the directions
taken by the curves. This was done in
1985 to calculate the composition of H2O
at the experimentally measured invariant
point where muscovite decomposes (in the
presence of sillimanite) at about 730oC
& 5 MPa). Though this particular
calculation was considered ill posed,
'H2O' was calculated to contain on about
36 to 41% H2O (by weight).

Bowen & Tuttle (on p.91 of GSA Memoir
74) wrote of this great flux of material
'If hydrous vapor, escaping from a
cooling granite intrusive, passes along
the contact the vapor at certain places
can selectively dissolve feldspars and
quartz and leave behind lime, magnesia,
P2O5, and perhaps iron.'

They considered it possible that the
'vapor' remained stationary, and the
flux was by diffusion through it. This
brings to mind the hypothesis of
Holdquist, and his genetic term
'venite'. Their alternative hypothesis
was that the material itself flowed,
bringing to mind the hypothesis of
Sederhold and his network of
'arterites'.

Were a second reaction to produce an
immiscible vapor, of differing
composition, whose properties changed
differently with pressure & temperature,
a variety of migmatitic phenomena could
be imagined.

The word 'vapor' brings to mind a single
gas, not a distinct collection of fluids
of differing compositions & densities,
possibly formed by differing reaction
mechanisms.

Just an idea to think about.


Bruce