1. ORIGIN OF MYRMEKITE AND METASOMATIC GRANITE
Lorence G. Collins
email: lorencec@sysmatrix.net
November 21, 1996; revised February 17, 1997
Two different generally-accepted hypotheses for the origin of myrmekite are examined in seven different photomicrographs.
Myrmekite, the vermicular intergrowth of quartz in plagioclase, is considered by some geologists to form by replacement of K-feldspar by Ca- and Na-bearing fluids.
KAlSi3O8 + Na+1 = NaAlSi3O8 + K+1
2KAlSi3O8 + Ca+2 = CaAl2Si2O8 + 4SiO2 + 2K+1
potassium feldspar myrmekite
This hypothesis seems logical, particularly when the K-feldspar crystal is relatively large (>1 cm) and when the myrmekite is tiny (less than 0.5 mm) and forms wartlike projections into the K-feldspar margins (as seen later in Figs. 2-5). But is the K-feldspar primary and is the myrmekite a secondary alteration? Does the above mass-for-mass equation actually represent the reaction in a rock where it is clear that the K-feldspar is primary and replaced by Na and Ca to form secondary plagioclase? Let's look at textures in a rock from Ausable Forks, New York (USA) (Fig. 1a and Fig. 1b).
If replacement of K-feldspar by plagioclase were mass-for-mass, as in the above balanced equations, tiny quartz vermicules should have formed in the secondary plagioclase in Fig. 1a and Fig. 1b, but none is present. Therefore, the replacement is not mass-for-mass but volume-for-volume. In the latter case, the greater density of secondary plagioclase (2.63) relative to the lesser density of K-feldspar (2.56) requires that silica cannot be released but is consumed in forming the secondary plagioclase that fills the space once occupied by the K-feldspar. Thus, although the balanced mass-for-mass equations appear to be quite logical, they cannot be used to explain the origin of myrmekite, at least in some rocks.
As further evidence that Ca- and Na-replacements of K-feldspar are not valid is shown by four different granites (Figs. 2-5), which have characteristics that make mass-for-mass reactions unlikely. If myrmekite in these granites were formed by Ca- and Na-bearing fluids that replaced K-feldspar, then the myrmekite in each host granite should be most abundant near fault zones, should be concentrated along fractures in K-feldspar crystals, and should totally replace the K-feldspar in some places, and none of these characteristics is found in the following examples. An exception that fulfills the aforesaid characteristics is illustrated in the fourth presentation.
A second hypothesis that is commonly used to explain the origin of wartlike myrmekite is to suggest that the K-feldspar is primary --- a high-temperature orthoclase, containing dissolved Na and Ca. At low temperatures the orthoclase, under stress, inverts to microcline and supposedly exsolves the Na and Ca to the margin of the crystal to form myrmekite. As above, balanced mass-for-mass equations are used to explain what happens. Because calcic plagioclase requires less silica in its lattice than in K-feldspar or sodic plagioclase, silica is left over to form quartz vermicules in the myrmekite proportional to the Ca-content.
KALSi3O8 = KAlSi3O8 + NaAlSi3O8 + 2SiO2
NaAlSi3O8 CaAl2Si2O8
CA(AlSi3O8)2
H-T K-feldspar K-feldspar myrmekite
Examples of photomicrographs of myrmekite with increasing sizes of quartz vermicules include Fig. 2, Fig. 3, Fig. 4, Fig. 5, and Fig. 6.
The examples in Figs. 2-5 have been purposefully selected to show textures in granites adjacent to wall rocks containing plagioclase having a range of An-contents from sodic to calcic. Other investigators have assumed that the K-feldspar in these granites is primary, formed at high temperatures, and that the myrmekite is a secondary alteration product, formed by exsolution. But if all these granites have essentially the same modal compositions or nearly so, then one has to ask: "Why do myrmekitic textures look so different for granites of the same modal composition?" Moreover, if the granite is an intrusive magmatic body, coming from some unknown source at depth, and the K-feldspar crystallized from this magma, how is it possible that the maximum sizes of quartz vermicules in the coexisting myrmekite in each granite correlate with increasing Ca-contents of the plagioclase in the adjacent wall rocks? Should not the composition of the K-feldspar in the granite be nearly independent of the wall rock? How would an intruding magma know what the wall rock composition is going to be?
In the granites illustrated in Fig. 2 (Cooma pluton but not the Cottonwood Creek pluton) and Fig. 3 and Fig. 4, I have observed all mineralogical changes between undeformed wall rocks through increasing degrees of deformation toward the associated granite. In early stages the first appearances of K-feldspar and myrmekite occur in the deformed wall rocks. Where the rocks are first deformed and where K-feldspar and myrmekite appear for the first time, the volume of myrmekite commonly far exceeds the volume of adjacent K-feldspar. The larger volume of myrmekite makes it impossible for the myrmekite to have formed by exsolution from the adjacent smaller K-feldspar crystal. Moreover, where the deformed parent rock is calcic diorite or gabbro, the Ca-content of the plagioclase in the myrmekite is greater than could have been possibly contained in the volume of the adjacent high-temperature orthoclase. Therefore, these relationships suggest that myrmekite is not formed by exsolution but by some other process.
The occurrences of muscovite in the two-mica granite (Fig. 5) and of other aluminous minerals (sillimanite, garnet, epidote, cordierite) in other granites give further support to the possible origin of such rocks by K-replacement processes, provided that thegranites are myrmekite-bearing. If these peraluminous granites have formed by replacement of wall rocks that contain plagioclase An45-100 in diorite or gabbro, this plagioclase is also relatively Al-rich in comparison to K-feldspar. Because displaced Al has low mobility, it tends to remain behind in a metasomatic granite to make it peraluminous.
Moreover, if calcic diorite and gabbro are the source rocks and they are strongly deformed in planar shear zones, then garnet, garnet-sillimanite, and garnet-sillimanite-cordierite gneisses could form instead of peraluminous granites, and such rocks would be improperly identified as metapelites.
If the former gabbro contains plagioclase near An100, as occurs in some places near Central City, Colorado, then the gneisses may have recrystallized products that are unrecognized as myrmekite. The quartz vermicules would be so thick that they would appear to be just large quartz grains in the ground mass.
Finally, it is quite apparent that reversing the direction of the two, aforesaid, balanced, mass-for-mass equations consumes quartz instead of producing it, so that myrmekite could not form in such a reaction. This obvious fact provides the logic for ridiculing any hypothesis which proposes that myrmekite forms by K-replacement of plagioclase. The wartlike projection of myrmekite into large K-feldspar grains (Figs. 2-5) further makes such an hypothesis seem silly. However, in the second presentation I show that secondary K-feldspar replaces primary plagioclase volume-for-volume and that myrmekite is formed simultaneously in the same process.
In the third presentation I show additional photos of field and microscopic relationships of myrmekite-bearing granitic rocks.
In a supplement, I provide 16 more photomicrographs to illustrate replacement textures. See link to article #43, below.
Dr. Lorence G. Collins Department of Geological Sciences California State University Northridge 18111 Nordhoff Street Northridge, California 91330-8266 FAX 818-677-2820