Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
UNITED STATES DEPARTMENT OF THE INTERIOR
GEOLOGICAL SURVEY
LAKE ELLEN KIMBERLITE, MICHIGAN, U.S.A.
by
E. S. McGee and B. C. Hearn, Jr.
U. S. Geological Survey
OPEN-FILE REPORT 83-156
This report is preliminary and has not been reviewed for conformity with U.S. Geological Survey editorial standards and stratigraphic nomenclature.
Any use of trade names is for descriptive purposes only and does not imply endorsement by the U.S.G.S.
1983
CONTENTS
Page
1 -- . -. I.*. _. .»._._-.,_, ._... ._. .l->l-)>l-) ±
Tn-f- pnHii ri~ i nn _______ ________ ............ _, _. 1J» I I W I w\JU W w I VI I X
Megacryst-xenocryst suite -- 2T 1 mon i *t" PC _-_- _, _-_- _. _- -.__..____ _. *5i I flfWil I w t *J ^
"i\ i
Pnmnnc i1~AV/ v/ 1 1 1 W v/ -> I wC
LI Vll Iv III'-* I U^ I Wl I ^
Garnet pyroxenite 6Spinel pyroxenite-peridotite 7
Temperature and pressure considerations 7Pnnp lucinnc ____ _._, -._ QV/v/llwlulOlv/lloJ ^ ^ ^ _> ^
AP!^ n nu/1 pHnmPnl" Q _-.___ __-.____ ___.._ Qfl^lX IIV/V*I^VdMIl|vllwO _ _ __ __ «. __ « mm mm mm m _> .__ __ . . . . . . i... __ . . __ . . . . . v j
References 10
ILLUSTRATIONS
Figure
1. Location map for the Lake Ellen kimberlite 122. Magnetic map of the Lake Ellen kimberlite, showing outcrops
P nrl npncnppl" n*it"c _. _- .- « « _ 1 ^UI|VJWIv/<_>W^wli*Wlv»O Aw
3. Fe203-MgTi03-FeTi03 composition of ilmenitesA. Megacrysts and xenocrysts, Lake Ellen kimberlite 14 B. Lake Ellen megacrysts and xenocrysts compared to ilmenites
from other North American kimberlites 144. 0303 vs. MgO of ilmenites 155. Ca-Mg-Fe composition of garnets 166. CaO vs. 0303 of xenocrysts, megacrysts, and xenolith garnets - 167. 0303 vs. Na£0 of clinopyroxenes 178. Thin section photos of eclogite and pyroxenite xenoliths
from Lake Ellen kimberlite - 189. Temperature-pressure diagram, Lake Ellen xenoliths, megacrysts,
f\ nrl ypnnp pvc1"C -.-.-.__-._-.__.____ _-.____ _______ 1QUI1\J /\\*»flV/X^|lf*Jw*J ______ _________ ______ _«. ._. . -_________ _ ...__ j, J
TABLES
Table
1. Megacrysts from the Lake Ellen kimberlite 202. Composite xenocrysts
A. Garnet-clinopyroxene 20B. Cl inopyroxene-orthopyroxene 20
3. Garnets from lithic inclusions 214. Pyroxenes from lithic inclusions 215. Oxides and kyanite from lithic inclusions 226. Feldspars from eclogite inclusions 22
THE LAKE ELLEN KIMBERLITE, MICHIGAN, U.S.A.
by E. S. McGee and B. C. Hearn, Jr.
ABSTRACT
The recently discovered Lake Ellen kimberlite, in northern Michigan, indicates that bedrock sources of diamonds found in glacial deposits in the Great Lakes area could lie within the northern U.S. Magnetic surveys show a main kimberlite 200 m in diameter and an adjacent body 25 x 90 m(?). The kimberlite cuts Proterozoic volcanic rocks that overlie Archean basement, but is post-Ordovician in age based on abundant Ordovician(?) dolomite inclusions. Xenocrysts and megacrysts are ilmenite (abundant, 12.5-19% MgO), pyrope- almandine and Cr-pyrope (up to 9.3% ^03), Cr-diopside (up to 4.5% ^03), olivine (Fo 91), enstatite and phlogopite. The kimberlite contains fragments of crustal schist and granulite, as well as disaggregated crystals and rare xenoliths of eclogites, garnet pyroxenites and garnet peridotites from a heterogeneous upper mantle. Eclogites, up to 3 cm size, show granoblastic equant or tabular textures and consist of jadeitic cpx (up to 8.4% Na20, 15.3% ^2^3)' pyrope-almandine, _+ rutile _+ kyanite _+ sanidine _+ sulfide. Garnet pyroxenite contains pyrope (0.44% 0^03) + cpx (0.85% Na20, 0.53% ^03) + Mg- Al spinel. Mineral compositions of rare composite xenocrysts of garnet + cpx are distinctively peridotitic, pyroxenitic or eclogitic. Calculated temperatures of equilibration are 920-1060 °C for the eclogites and 820-910 °C for the garnet pyroxenite using the Ellis-Green method. Five peridotite garnet-clino- pyroxene composite xenocrysts have calculated temperatures of 980-1120 °C using the Lindsley-Dixon 20 kb solvus. Spinel pyroxenite and clinopyroxene- orthopyroxene composites have lower calculated temperatures of 735 °C and 820-900 °C, respectively. Kyanite-bearing eclogites must have formed at pressures greater than 18-20 kb. Using the present shield geotherm with a heat flow value of 44mW/m^ for the time of kimberlite emplacement, the eclogite temperatures imply pressures of 35-48 kb (105-140 km) and the garnet pyroxenite temperatures indicate pressures of 24-29 kb (75-90 km). Temperatures of two peridotitic garnet-cpx composite xenocrysts if on a shield geotherm, imply pressures within the diamond stability field.
INTRODUCTION
The Lake Ellen kimberlite is located in the upper peninsula of Michigan, near the Wisconsin-Michigan border (Fig. 1). It is about 15 km northeast of Crystal Falls, Michigan and about 1.5 km west of Lake Ellen. The kimberlite was first recognized about 15 years ago by an exploration geologist. A recent description of the kimberlite, its geologic setting, and magnetic and gravity traverses (Cannon and Mudrey, 1981) emphasizes its significance as indicative of possible bedrock sources of the widespread diamond finds in alluvial and glacial deposits in the Great Lakes area. Cannon and Mudrey (1981) also summarize other occurrences of localized subsidence or intense deformation in
the region that may be related to emplacement of kimberlite or to other diatreme- forming processes. In the early 1950's, a ground magnetometer survey found erratic high values over part of the kimberlite, but there were no exposures of kimberlite at that time and the erratic values were ascribed to the common problem of boulders of iron formation in the glacial deposits (Gain and Wier, 1956).
The kimberlite cuts across lower Proterozoic X metavolcanic rocks, southeast of the Amasa dome of Archean gneiss. The age of the kimberlite has not yet been determined, but it is post-Ordovician as it contains abundant inclusions of fossiliferous Ordovician(?) dolomite similar to the Black River Group. The nearest exposure of the Black River Group is now about 60 km away, and the Black River Group would have been at least 200 m strati graphically above the present surface when the kimberlite was intruded (Cannon and Mudrey, 1981).
Most of the area is covered by glacial till and alluvium, and the kim berlite has only four outcrops, three of which are in a logging road. The largest outcrop is 2 by 4 m in size, and like the others, consists mainly of small fragments of altered kimberlite.
A ground magnetometer survey was done by the authors July 12-15 and August 12, 1980, and August 9, 1981 using a Geometries model G-812/826 proton magnetometer. A grid of stations at intervals of 50 feet (15 m) and locally 25 feet (7.5 m) or less, was marked by measuring distances with a tape along lines established using a Brunton compass. East-west lines 100 feet (30.5 m) apart were surveyed on each side of a central north-south line. Additional stations halfway between those east-west lines were located by tape measurement from each line. Readings every 1-1.5 hours at a primary base station 370 feet (115 m) south of the kimberlite, or at a secondary base station, were used to make linear corrections for diurnal variations. Magnetic values are believed to be accurate to +_ 3 gammas. Two pipes have been identified and contoured (fig. 2) from the survey. The main pipe is roughly circular and about 200 m in diameter. A smaller, satellitic body to the northeast is less well defined, but is about 25 by 90 m in size.
The kimberlite in the outcrops is mainly a rubbly, partially clayey material with many xenocrysts, lithic inclusions from shallow levels, and a few xenoliths of deeper origin. Several shallow pits were dug, and in a few, hard kimberlite was reached. The hard kimberlite is extensively serpentinized, and contains xenocrysts and xenoliths like those found in the loose rubble of the outcrops.
MEGACRYST-XENOCRYST SUITE
The suite of megacrysts (greater than 1 cm) are distinguished from the suite of xenocrysts (less than 1 cm) by their grain size. Megacrysts and xenocrysts which are contained in the kimberlite have been primarily selected from panned concentrates and from bag samples of loose material. Xenocrysts average 2-4 mm in size. The megacryst suite is composed of ilmenite, clinopyroxene, and garnet. Representative analyses of each are shown in Table 1. The xenocryst suite is more diverse, being composed of: ilmenite, garnet, clinopyroxene, orthopyroxene, phlogopite, and olivine. In addition, several
composite xenocrysts of garnet-clinopyroxene and clinopyroxene-orthopyroxene have been found. These composites are equant mineral pairs, l-2mm in size, which are embedded in each other as coexisting grains and are not intergrowths.
ILMENITES
The ilmenites analyzed thus far are mainly megacrysts. The megacrysts range in size from 1-3 cm, with most falling in the range 1-1.5 cm. The megacrysts are both multigranular and single grains. Many of the ilmenite megacrysts have a single grain 'core' surrounded by a multigranular rim. Four xenocrysts from panned concentrate (2-4 mm) and 5 xenocrysts (1 mm) picked from the kimberlite matrix have been analyzed. The MgO content in megacrysts and xenocrysts is high, ranging from 12.54 to 19.10 wt. %. The ilmenite grains have 45-60% geikelite component and 2-7% hematite component (Fig. 3A). This hematite component is slightly lower than other ilmenites from North American kimberlites (Fig. 3B). There is no distinct difference between Lake Ellen xenocryst and megacryst compositions, indicating that they may have been derived from the same region at depth. 0^03 - MgO compositions (Fig. 4) show a wide range of 0203 contents (0.35-1.52 wt. %), and are more MgO rich than the ilmenites used by Haggerty (1975) in constructing the ilmenite parabola. Again, the ilmenite xenocrysts and megacrysts are not compositionally distinct. The MgO-rich, and Fe203-poor nature of these ilmenites indicates that they formed under reducing conditions and have not undergone extensive oxidation.
GARNETS
Garnets from Lake Ellen are predominantly found as xenocrysts. These xenocrysts encompass a range of colors and have been subdivided on the basis of color to identify any relationships or trends. The techniques of sorting and constraints on identifying the colors of the garnets were the same as those discussed by Hearn and McGee (1982). The color groupings used were: purple, red, pink, orange, red-orange, and light orange. The purple, red, and pink garnet groups tend to be more Mg-and Cr-rich than the three orange garnet groups (Figs 5 & 6). The orange, red-orange, and light orange garnets are close in Ca-Mg-Fe composition to the garnets from eclogite xenoliths (Fig. 5) and many are probably fragments from disaggregated eclogite inclusions. All of the purple, red, and pink garnets fall within the Ca-Mg-Fe field of garnets from Montana garnet peridotites (fig. 5) which is similar to peridotite garnets from other kimberlites. The purple, red, and pink garnets also are within or close to the CaO-Cr203 field of Iherzolitic garnets (Sobolev, et. al, 1973) (Fig 6). This, combined with the higher Mg content (Fig. 5) indicates a peridotitic affinity for the purple, red, and pink garnets. No Cr-rich, Ca-poor garnets typical of garnets from diamond-bearing harzburgites (Boyd and Finnerty, 1980) have been found at Lake Ellen, perhaps because of the somewhat limited sampling of garnet xenocrysts and megacrysts.
Garnet megacrysts, not as abundant as xenocrysts, have been found in the loose rubbly material of the outcrops and in chunks of the hard kimberlite, dug from shallow pits. The megacryst analyzed so far is more similar to the peridotitic than to the eclogitic garnets (Figs. 5 and 6).
PYROXENES
The xenocryst suite from Lake Ellen includes abundant clinopyroxene and rare orthopyroxene. Several clinopyroxene megacrysts have also been analyzed (Table 1). The ^0 and 0203 contents of the megacryst and xenocryst clino- pyroxenes can be used to distinguish associated xenolith types. The clinopy- roxenes are eclogitic, pyroxenitic, and peridotitic in affinity (Fig. 7). Clinopyroxenes from eclogites have ^0 contents which range from 4-8 wt. % and low 0203 (less than 0.5 wt. %). Peridotitic Clinopyroxenes have lower Na20 (less than 5 wt. %) and high 0^03 contents (0.5-4.5 wt. %). Pyroxenitic Clinopyroxenes have low Na20 (less than 1 wt. %) and low 0203 ( about 0.5 wt. %) which respectively distinguishes them from eclogitic and peridotitic Clino pyroxenes. Two groups of clinopyroxene xenocrysts, shown as dashed-line fields in Fig. 7, are similar, particularly in 0^03 content, to the Clinopyroxenes from peridotites or to the Clinopyroxenes from pyroxenites and megacrysts.
Orthopyroxene xenocrysts are rare compared to the Clinopyroxenes. They contain about 4 wt. % Al203 and about 0.5 wt. % CaO. One orthopyroxene xenocryst appears anomalous because it contains 1.2 wt. % A1203 and 1.5 wt. % CaO, which is unlike any other orthopyroxene analyzed thus far from this kimberlite. The high CaO content indicates a high temperature of crystalli zation. All of the orthopyroxenes contain approximately 1 wt. % Cr203, sug gesting disaggregation from a Cr-rich assemblage similar to that of the Cr- rich composite garnet-clinopyroxene xenocrysts.
OLIVINES
Most of the olivines in the kimberlite have been serpentinized, but a few fresh olivines have been found in kimberlite matrix and in panned concentrates. The olivines have a narrow range of forsterite content from Fogo.5 t° Fogi.o* and contain an average of 0.38 wt. % NiO, suggesting that they are xenocrysts rather than phenocrysts (Dawson, 1980). They show no zonation from core to rim.
COMPOSITE XENOCRYSTS
The composite xenocryst suite is composed of two types of pairs: clinopy- roxene-garnet and clinopyroxene-orthopyroxene. Although the grain size of the members of the pairs is rarely equal, they appear to be in equilibrium with one another. They are most likely grains from disaggregated xenoliths because 1) there is no alteration or kimberlite matrix between the grains, 2) they have an equant rather than intergrowth texture, and 3) neither is totally surrounded by the other as an inclusion. The compositions (Table 2) of the composite xenocrysts indicate derivation from eclogites, garnet pyroxenites, and peridotites.
In the garnet-clinopyroxene composites the color of the garnet grains was used as a guide to the xenolith type: purple garnets for peridotitic composites, and orange garnets for eclogitic composites. A greater number of peridotitic pairs were selected for analysis because garnet peridotite xenoliths have not yet been found and the composite grains have been used to represent a 'missing portion' of the xenolith population. As with the xenocrysts, the
chromium content of both the garnet and clinopyroxene, and the sodium in the clinopyroxene are characteristic of the rock type from which they were derived. The clinopyroxenes from the garnet-clinopyroxene composites are higher in Na20 and 0303 than the clinopyroxene xenocrysts which have been analyzed (Fig. 7). This is most likely a result of limited or biased selec tion of the xenocrysts. Four of the five clinopyroxenes from composites are also distinctly more Cr-rich and Na-rich than clinopyroxenes from Montana garnet peridotites (Fig. 7) (Hearn and McGee, 1983), indicating an unusually Cr-Na-rich bulk composition for garnet peridotites represented in the Lake Ellen kimberlite.
The clinopyroxene-orthopyroxene pairs are peridotitic in nature with the clinopyroxenes containing about 1 wt. % Na20 and 1 wt. % 0303. The orthopyroxenes are very similar in composition to most of the orthopyroxenes found as xenocrysts (A1203 = 4 wt. %; CaO = 0.5 wt. %). On a 0^03 vs Na20 plot (Fig. 7) the clino pyroxenes from cpx-opx composites fall within the field of clinopyroxenes from Montana garnet peridotites, unlike the clinopyroxenes from the clinopyroxene- garnet pairs which are much richer in both Na20 and Cr^. These clinopyroxenes indicate two groups of peridotites at the Lake Ellen kimberlite.
The composite xenocrysts have been included with the xenoliths in plots and in temperature-pressure calculations. Although the composite xenocrysts give limited information compared to the complete mineral assemblage in xenoliths, the composite xenocrysts represent rock types which must also be present in the upper mantle.
LITHIC INCLUSIONS
Lithic inclusions in the Lake Ellen kimberlite consist of both crustal and upper mantle rock types. The crustal inclusion suite predominates and consists of dolomite, phyllite, amphibolite, altered gabbro, and a few granulites. This study has concentrated on the less abundant upper mantle inclusion suite consisting of eclogite, garnet pyroxenite, a spinel pyroxenite- peridotite, and a serpentinized peridotite inclusion containing clinopyroxene as the only remaining fresh phase. Representative analyses of phases are given in Tables 3,4,5, and 6.
ECLOGITE INCLUSIONS
The eclogite suite is the largest group of xenoliths collected from Lake Ellen, comprising about 85% of all the upper mantle xenoliths collected. The eclogites range in size from 0.5 cm fragments of inclusions to 3.0 cm inclusions. They have a granulitic texture, and most show a distinctive alteration of the clinopyroxenes. The eclogite assemblage consists of garnet + clinopyroxene +_ rutile +_ kyanite _+ sanidine +_ corundum +_ sulfide. The majority of eclogite samples contain only garnet, clinopyroxene, and rutile. The compositions of the garnets and clinopyroxenes in the eclogites are not significantly different when kyanite or sanidine are also present.
The garnets in the eclogites are generally orange, contain more iron than magnesium, and are widely scattered near the central part of a Ca-Mg-Fe
ternary (Fig. 5), showing a range of 30% Fe component. This lack of clus tering of garnet compositions from eclogites has been observed in other kim- berlites as well (Meyer, 1977). The chromium content of the garnets is low (about 0.1 wt. %).
The clinopyroxenes in the eclogites, are mostly altered, appearing cloudy light green or almost white in hand specimen, and showing a very fine network of alteration in thin section (Fig. 8A). In most samples the grains still have unaltered cores which are visible in thin section and can be analyzed. The clinopyroxenes are very rich in sodium and aluminum. The sodium content ranges from 3-8 wt. % with most falling between 6 and 7 wt. % (Fig. 7). The aluminum content parallels the sodium, ranging from 10-16 wt. % with the average around 14 wt. % Al303. Chromium content of eclogitic clinopyroxenes is low, rarely reaching 0.10 wt. % (Fig. 7).
Rutile ranges in size from grains nearly 1.5 mm long to tiny needles included in clinopyroxenes. The rutile is red-brown to yellow-brown and is fairly pure although some contain as much as 1.5 wt. % FeO (Table 5). Kyanite occurs as small (1 mm), bladed grains which are very pale blue to white. In one sample (KB2-3), the kyanite has a preferred orientation (Fig. 8B). Kyanite contains approximately 0.3 wt. % FeO and up to 0.2 wt. % C p2°3 (T3^ 6 5). Corundum has been found in only two samples so far, both of which also contain kyanite. The corundum occurs as small (0.05-0.6 mm) blades near kyanite and may be part of a reaction occurring during kimber- lite ascent. It contains only minor amounts of ^303 (0.2 wt. %), Ti02 (0.2 wt. %), and FeO (0.3 wt. %). Sulfides occur as rounded grains (up to 0.8 mm size) and as smaller patches within the alteration between pyroxenes. They have not yet been analyzed but preliminary observations indicate that they contain more than one phase.
Potassium feldspar is present in 5 of the eclogites studied. The grains are up to 1.8 mm in size. A few of the feldspars show twinning, and a few also have minor alteration along the edges. The feldspar is very potassium- rich, containing 14-16 wt. % K£0 and only 0.3-1.3 wt. % Na20 (Table 6). Barium content of these feldspars is significant; in four samples BaO is about 0.5 wt. % and in one it is about 2.0 wt. % (Table 6). Within each eclogite the barium in the feldspars is constant from grain to grain, and it does not appear to be zoned within the grains. The small grain size of the feldspars has precluded study by x-ray diffraction to determine the structural state. Based on the findings of other workers (Smyth and Hatton, 1977; Prinz, et al, 1975) and on the high calculated temperatures from garnet-pyroxene pairs, the feldspar is probably sanidine rather than orthoclase.
GARNET PYROXENITE
Garnet pyroxenite (1.5 cm size) has a granulitic texture and is partially serpentinized (Fig. 8C). It consists of garnet + clinopyroxene + spinel + sulfide. Although the garnet is slightly richer in MgO than many of the eclogitic garnets, it is not clearly distinguishable from them. However, the pyroxene can be used to distinguish the garnet pyroxenite and eclogite xenoliths because sodium is 0.8 wt. % and aluminum is 4 wt. % in the garnet pyroxenitic clinopyroxenes, much lower than they are in the eclogitic
clinopyroxenes (Fig. 7; Table 4). The low chromium content of the pyroxene in garnet pyroxenite, 0.5 wt. %, distinguishes it from peridotitic clino- pyroxene. The spinel in the garnet pyroxenite is red-brown to dark brown, has a grain size of 0.5 mm, and is nearly surrounded by garnet in the sample (Fig. 8C). It is Al-rich and Cr-poor, containing 55.7 wt. % A^OS and 9.7 wt. % 0303 (Table 5).
SPINEL PYROXENITE-PERIDOTITE
One small (about 1.5 cm) sample has a fine-grained (0.25-1 mm) tabular to granular texture and is extensively altered to serpentine in the olivine rich part, giving it a banded appearance (Fig. 8D). Two thirds of the sample con tains altered olivine, spinel, and clinopyroxene and was probably peridotitic when fresh. The remaining third of the sample is pyroxenitic containing spinel, clinopyroxene, orthopyroxene and a few fresh olivines. Black spinel occurs throughout the section, and shows some tendency to be concentrated in layers (Fig. 8D). It is an aluminum-rich spinel with about 8.5 wt. % 0203, (Table 5) and slightly more iron than the spinel in the garnet pyroxenite. The clino pyroxene is fresh and its 0203 and Na20 contents are close to some of the clinopyroxene megacrysts and most of the clinopyroxene xenocrysts (Fig. 7). It contains about 1 wt. % Na20, 4.5 wt. % A1203, and 0.35 wt. % 0203 (Table 4). The orthopyroxene is less abundant than the clinopyroxene and contains slightly less A1203 and CaO (3 and 0.2 wt. %, respectively) than most of the orthopyroxenes found as xenocrysts or as members of composite grains (Tables 2B,4).
TEMPERATURE AND PRESSURE CONSIDERATIONS
Temperatures were calculated for the xenoliths and composite xenocrysts using A.A. Finnerty's TEMPEST program. The Lindsley and Dixon (1976) 20 kb clinopyroxene thermometer was used for the peridotitic garnet-clinopyroxene composite xenocrysts. However, the Lindsley and Dixon thermometer could not be used for the other xenolith types because of the high Ca/(Ca+Mg) values of the clinopyroxenes. The Ell is and Green (1979) 30 kb garnet-clinopyroxene thermometer was used for the eclogites and garnet pyroxenites. The Wells (1977) thermometer, which does not require garnet and extends to lower temper atures than the Lindsley-Dixon thermometer, was used for clinopyroxene-ortho- pyroxene composite xenocrysts and for the spinel pyroxenite-peridotite. The Boyd and Nixon (1973) calcium in orthopyroxene thermometer was used for one high calcium orthopyroxene xenocryst.
Pressures can be directly determined only for those samples which contain orthopyroxene. The MacGregor (1974) barometer, which uses Al203 in enstatite, indicated pressure of 12.3 kb for the spinel pyroxenite-peridotite, and a pressure range of 13 to 17 kb for the clinopyroxene-orthopyroxene composite xenocrysts. Both are within the spinel stability field. However, because there is disagreement about the direction of slope of the A^O^ isopleths in the spinel stability field (MacGregor, 1974; Obata, 1976; Dixon and Presnall, 1977), the pressures estimated using MacGregor's (1974) barometer may be incorrect or controversial. Although the pressures are useful for indicating derivation from the spinel stability field, rather than from the garnet
stability field, the pressures were not used in constructing a pressure- temperature diagram for inclusions in the Lake Ellen kimberlite. Only the pressure calculated for the high calcium orthopyroxene xenocryst was used, as it indicated derivation from the garnet stability field.
Calculated temperatures have been projected onto the present day shield geotherm of 44 mW/m'2 for northern Michigan (Fig. 9). Assumption of a shield geothermal gradient is based on the lack of igneous activity in northern Michigan after Keweenawan volcanism in the late Proterozoic, except for post- Ordovician emplacement of kimberlite.
Highest equilibration temperatures were calculated for a clinopyroxene from an altered peridotite (1292 °C) and for a high calcium orthopyroxene xenocryst (1350 °C). High equilibration temperatures (980-1120 °C) were also calculated for composite xenocrysts of garnet peridotitic affinity. The eclogites equilibrated at intermediate temperatures (920-1060 °C), and the garnet pyroxenites have the lowest range of calculated temperatures (820-910 °C) for garnet-bearing samples. The clinopyroxene-orthopyroxene composites have a temperature range (820-900 °C), which is similar to that of the garnet pyroxenite samples. The lowest equilibration temperature was calculated for the spinel pyroxenite-peridotite (735 °C).
Fields for garnet peridotites from Montana and garnet peridotites and garnet pyroxenites from Colorado-Wyoming are shown for reference in Fig. 9. Two of the composite garnet-clinopyroxene pairs fall within the field of garnet peridotites from Colorado-Wyoming. Without an independent pressure determination for peridotitic xenocrysts, the presence of an "inflected geotherm" of high temperature-pressure peridotites in northern Michigan is uncertain. Two pyroxenes possibly indicate an inflected geotherm. The clinopyroxene from an altered peridotite and the high calcium orthopyroxene xenocryst are similar in temperature to the high temperature peridotites from Montana and South Africa.
Within the eclogite group, the kyanite and sanidine bearing samples are intermingled with other eclogites throughout the temperature range covered. Based on only four samples, there is an apparent trend of increasing temper ature from kyanite eclogite (928 °C) to kyanite-sanidine eclogite (984 °C) to sanidine eclogites (1044, 1058 °C). This trend suggests increasing potassium and decreasing aluminum content with depth, indicating derivation of the eclogites from a layered upper mantle.
Calculation of pressure remains a problem with all of these samples. The presence of kyanite in the eclogite samples indicates pressures were at least 18-20kb (Green, 1969). Projecting the calculated temperatures onto the shield geotherm is reasonable if the geotherm is the same today as it was at the time of kimberlite emplacement. If the MacGregor (1974) spinel facies pressures are correct, the spinel pyroxenite-peridotite and the clino- pyroxene-orthopyroxene composite xenocrysts indicate that the geotherm may have been higher during emplacement. Without reliable pressure calculations and with only one high temperature-pressure point the position and shape of the emplacement geotherm cannot be adequately described.
The relationship of the Lake Ellen samples to the diamond-graphite stability curve is of interest, particularly because diamonds have been found in the glacial drift of Wisconsin, Michigan, Illinois, Indiana, and Ohio. If the present day geotherm is a realistic estimate of the tempera tures and pressures of origin, some of the garnet-clinopyroxene pairs and some eclogites indicate derivation from within the diamond stability field (Fig. 9). This could mean that the kimberlite could be a source for some of the diamonds found in glacial drift in southern Wisconsin, but only if carbon were present at depth, below about 120-135 km, and if diamond were preserved during kimberlite ascent. A further constraint to note, is that if the emplacement geotherm was 100 to 150 degrees higher than the present day geo therm, the highest temperature composite pairs and eclogites would fall within the graphite stability field. So far, no diamonds have been found at this kimberlite, and more conclusive evidence is needed to evaluate the probability of their occurrence here.
The ranges of temperature for eclogites and garnet peridotitic composite xenocrysts overlap. Two samples of garnet pyroxenite and several clinopyroxene- orthopyroxene composite xenocrysts occupy a temperature range just below the lowest temperature eclogite, and the lowest temperature of 750 °C is given by a spinel pyroxenite-peridotite. Assuming that the temperatures are directly related to depth, the upper mantle section consisted of intermixed eclogite and garnet-peridotite overlain by garnet pyroxenite and spinel peridotite. Potassium was high enough to form feldspar in eclogites from a range of depths from at least 120 to 140 km, suggesting that the upper mantle in that depth range was enriched in potassium and probably in other elements easily removed by small degrees of partial melting. Thus the 120-140 km depth range of the upper mantle had not undergone partial melting after eclogite formation.
CONCLUSIONS
The Lake Ellen kimberlite, although poorly exposed, contains samples which reveal characteristics of a heterogeneous upper mantle. A wide range of xenocryst and xenolith types and compositions indicate that the kimberlite has sampled eclogite, garnet pyroxenite, spinel pyroxenite-peridotite, and garnet peridotite similar to upper mantle sections in other portions of the United States. Calculated temperatures for the inclusions show that they represent a range of equilibration temperatures of 700-1300 °C, and thus were derived from depths from 55 to 160 km or more assuming a shield geotherm. Some of the higher temperature samples indicate that the kimberlite could contain diamond.
ACKNOWLEDGMENTS
We thank W. F. Cannon for bringing the kimberlite to our attention, for providing the initial samples, and for a field trip to the kimberlite. The authors thank J. J. McGee, C. M. Sears, and W. L. Hearn for assistance in magnetic surveying and sample collection. We also thank A. A. Finnerty for giving us a version of his TEMPEST computer program for calculating temperatures and pressures of inclusions.
9
References
Agee, J. J., Garrison, J. R., Jr., and Taylor, L. A. (1982) Petrogenesis of oxide minerals in kimberlite, Elliot County, Kentucky. American Mineral ogist, v. 67, p. 28-42.
Bence, A. E. and Albee, A. L. (1968) Empirical correction factors for the electron microanalysis of silicates and oxides. Journal of Geology, v. 76, p. 382-403.
Boyd, F. R. and Finnerty A. A. (1980) Conditions of origin of natural diamonds of peridotitic affinity. Journal of Geophysical Research, v. 85, p. 6911- 6918.
Boyd, F. R. and Nixon, P. H. (1973) Origin of the ilmenite-silicate nodules in kimberlites from Lesotho and South Africa: in Nixon, P. H., ed., Lesotho Kimberlites, Lesotho National Development Corp., Maseru, p. 254-268.
Bundy, F. P., Bovenkerk, H. P., Strong, H. M., and Wentorf, R. H., Jr. (1961) Diamond-graphite equilibrium line from growth and graphitization of diamond. Journal of Chemical Physics, v. 35, p. 383-391.
Cannon, W. F. and Mudrey, M. G., Jr. (1981) The potential for diamond-bearing kimberlite in Northern Michigan and Wisconsin. U. S. Geological Survey Circular 842.
Dawson, J. B. (1980) Kimberlites and their xenoliths. Springer-Verlag, New York, 252 p.
Dixon, J. R. and Presnall, D. C. (1977) Geothermometry and geobarometry ofspinel Iherzolite in the system CaO-MgO-Si02» Extended abstracts, SecondInternational Kimberlite Conference, Santa Fe, New Mexico; unpaginated.
Eggler, D.H., McCallum, M. E., and Smith, C. B. (1979) Megacryst assemblages in kimberlite from northern Colorado and southern Wyoming: Petrology, geothermometry-barometry, and areal distribution: j£ Boyd, F. R., and Meyer, H. 0. A., eds., The Mantle Sample: Inclusions in Kimberlites and Other Volcanics; Proceedings Second International Kimberlite Conference, v. 2, American Geophysical Union, p. 213-226.
Ellis, D. J. and Green, D. H. (1979) An experimental study of the effect of Ca upon garnet-clinopyroxene Fe-Mg exchange equilibria. Contributions to Mineralogy and Petrology, v. 71, p. 13-22.
Gair, J. E. and Wier, K. L. (1956) Geology of the Kiernan quadrangle, Iron County, Michigan: U. S. Geological Survey Bulletin 1044,
Green, T. H. (1969) The diopside-kyanite join at high pressures and temperatures. Lithos, v. 2, p. 334-341.
Haggerty, S. E. (1975) The chemistry and genesis of opaque minerals in kimberlites. Physics and Chemistry of the Earth, v. 9, p. 295-307.
10
Hearn, B. C., Jr. and McGee, E. S. (1982) Garnets in Montana diatremes: a key to prospecting for kimberlites. U. S. Geological Survey Open-File Report 82-722, 43 p.
Hearn, B. C., Jr. and McGee, E. S. (1983) Garnet peridotites from Williams kimberlite, North-Central Montana, U.S.A. U.S. Geological Survey Open File Report 83-172, 25 p.
Kennedy, C. S. and Kennedy, G. C. (1976) The equilibrium boundary between graphite and diamond. Journal Geophysical Research, v. 81, p. 2467-2470.
Lindsley, D. H. and Dixon, S. A. (1976) Diopside-enstatite equilibria at 850° to 1400°C, 5 to 35 kb. American Journal of Science, v. 276, p. 1285-1301.
MacGregor, I. D. (1974) The system MgO-Al 2 03-Si02: solubility of A1 2 03 in enstatite for spinel and garnet peridotite compositions. American Mineralogist, v. 59, p. 110-119.
McCallum, M. E., Eggler, D.H., and Burns, L. K. (1975) Kimberlitic diatremes in northern Colorado and southern Wyoming. Physics and Chemistry of the Earth, v. 9, p. 149-162.
Meyer, H. 0. A. (1977) Mineralogy of the upper mantle: a review of the minerals in mantle xenoliths from kimberlite. Earth-Science Reviews, v. 13, p. 251-281.
Mitchell, R. H. (1979) Mineralogy of the Tunraq kimberlite, Somerset Island, N.W.T., Canada. _j_n_ Boyd, F. R. and Meyer, H. 0. A. eds.: Kimberlites, Diatremes, and Diamonds: Their Geology, Petrology, and Geochemistry; Proceedings Second International Kimberlite Conference, v. 1, American Geophysical Union, p. 161-171.
Obata, M. (1976) The solubility of Al 2^3 in orthopyroxenes in spinel and plagioclase peridotites and spinel pyroxenite. American Mineralogist, v. 61, p. 804-816.
Prinz, M. A., Manson, D. V., Hlava, P. F., and Keil, K. (1975). Inclusions in diamonds: garnet Iherzolite and eclogite assemblages. Physics and Chemistry of the Earth, v. 9, p. 797-815.
Smyth, J. R. and Hatton, C. J. (1977) A coesite-sanidine grospydite from the Roberts Victor kimberlite. Earth and Planetary Science Letters, v. 34, p. 284-290.
Sobolev, N. V., Laurent'ev, Yu G., Pokhilenko, N. P., and Usova, L. V. (1973) Chrome-rich garnets from the kimberlites of Yakutia and their parageneses. Contributions to Mineralogy and Petrology, v. 40, p. 39-52.
Wells, P. R. A. (1977) Pyroxene thermometry in simple and complex systems. Contributions to Mineralogy and Petrology, v. 62, p. 129-139.
11
48'N--
Kimberlite /
90* W88°W
Figure 1. Location map for the Lake Ellen Kimberlite
1 9
LAKE
ELL
EN K
IMBE
RLIT
E^
- 10
0 y
MAGN
ETIC
CON
TOUR
S (-
58,0
00)
..-«
- 50
y M
AGNE
TIC
CONT
OURS
GLA
CIAL
AND
ALL
UVIA
L DE
POSI
TS
INFE
RRED
KIM
BERL
ITE
CONT
ACT
KIM
BERL
ITE
OUTC
ROP
KIM
BERL
ITE
PRO'
SPEC
T PI
T
.HEM
LOCK
FM
OUT
CROP
(PRO
TER0
20IC
X)
100
200
Met
ers
I
Fig
ure
2.
Mag
netic
map
of
the
Lake
Elle
n
kim
berl
ite,
show
ing
outc
rops
an
d pr
ospe
ct
pit
s.
Con
tour
s ar
e to
tal
inte
nsi
ty
in
gam
mas
min
us 5
8,00
0 ga
mm
as.
MgT
103
Lake
E
llen
meg
acry
sts
Lake
Elle
n x
en
ocr
ysts
*. E
llio
t C
o.,
Ken
tuck
y m
egac
ryst
s (A
gee
et
al ,
1982
)
-" ,
Tun
raq,
S
omer
set
Isla
nd
C
anad
a;--
' xe
nocr
ysts
(M
ltch
ell,
19
79)
Col
orad
o-W
yom
ing
meg
acry
sts
(McC
allu
m e
t al
, 19
75)
Fig
ure
3.
F
e203
-MgT
i03-
FeT
i03
tern
ary
, ilm
enites.
A)
Exp
ande
d sc
ale
sh
owin
g m
egac
ryst
an
d xe
no-
cry
st
ilme
nite
s fr
om
Lake
E
llen
. B)
F
ield
fo
r La
ke
Elle
n m
egac
ryst
s an
d xe
nocr
ysts
com
pare
d to
ilm
en
ites
from
oth
er
Nor
th
Am
eric
an
kim
be
rlite
s
"/
A
- /
-\-
/ \
7^Y
/Av/A
Y.//A
'.V
/AV
/AV
.V,
LAKE
EL
LEN
IL
ME
NIT
ES
^ £
> M
egac
ryst
s ^O
Xen°
cr%ysts
(2
-4m
m)
t /t-
sin
gle
cry
sta
l ' m
ul
tig
ran
ul a
r O
X
enoc
ryst
s fr
om kim
be
rlite
m
atr
ix
(Fe2
03) B\
A
10
(Mg
Ti0
3)
fin50
4540
3530
4.0QMegacrysts (>lcm)
(l-4mm)
4 6 8 10 12MgO in weight percent _
Figure 4. Wt. % 01703 vs wt. % MgO of 1lmen1tes from the Lake Ellen klmberlite.
15
LaO
in
w
eig
ht
pe
rce
nt
_*
%>
-._M
: X
.,°>
c
l <i%
^
<*,
--.«4^ ^ v^ -t-
A"O
\*
VO
in
mo>
o
3 «
-*
O
O-U
D
3
rt-
10
n>IV o o st
n> * 1
^rt
U3
(OO
l -»
n>3
O
IV
O<-»
«
<ox»
^*
<
rt-
3"
IB 3 r*
IV
O
x-o
3
On>
o
ua 3
O>
T3
0
0 »
(^
*<
-*
w» r*
H
- w X IV 3 8 C2 (/> r* to
OD
O
130 -0
-O
Z
n>
- c
mC
X 3
-J
-
7C
-O (^
'
<itx>
r- O
30
-"-»
<»
<0
0>
Q
u3-
3
1ri-io O
i n
»0
»
»
3
n3
^
<o n
*^^
"""^ x o
i
3 0 O "< (/>
rt-
I/)
5
c o
4 o> C
L
sz
o> s c 3
CO
o CVJ
2 1 -
1_ 1
1
_ i
-4
-1 h
_ i
f
-T
4
_ i_
_.
i '
i
rri±
n
"7p
"
( | |
;
-
nr.^
-- .
_ .. r
-
-TI^
4
_^ __ !
_^_i 1 '
- i (-1
- j
t
1 1 ,
--4-4
-
; |
. '
i
_^ i .
^i
1 L
_ i,
i.-
.^_
; .
~ field
o
f cp
x M
onta
na
garn
et
_L_t
pe
rid
otit
es
- - !
,Z
i
D
1 j i j
,
i \ ,
}f
M^
f ^«
L!__JL k
|s i -f U KM it (., i dF
ifl
/ J1r - « s ^ t^ y -1 t* V ? f/
tig i
/ / ^ f
in
i/ i r K
L L/
'^ /s i
i -^
I
-4
\
i i i L
i
i t
I-
~, ...
. + --
i '
17 -
"1 72 K
-
^J \ !
-
s -.
k^
i i
i
1
i
1
1 I
i
-1
^
1 1
-4
_ ;
,
J"^
4~
,
-^^*
.^
i
_ V
i \
j-
/T
- t - i- i t- > / -
^
"*" 2
. L.:. __ .
~ 1
>.
.
(.
_L
, __
_
^ .
_ r _
,.. _
i
_4
_f.
__- j.
t- i. --r -
t 1 (_
.
_(-^
^ .
sd i i -j i itl
/!
""
, '
"^
*~ J
-J, V
f~j
1 i . -j i i
-^ J
- -;
,
< __
!_
_^
:
..::
-_
._,
.. ...
..
::.;:;
;.:
.....
....
. .
. i
. .
. .
.
' .
.. .
. 1
.
....
....
-
Y
.... ,._
(..
_
.
h r:
'.-L
-i.-f.4
-.
1 '
(
[ 1 ~"~
------
L 6
3
.T.
. .
^
. .
,
1~T
! '
' '
\
-
A
3 4
-
-"
i i f t
- -
(
-
i
...
. ....
..j. ...
....
.
1.
.4 ..
.. j 1 1 f
. -,
4. J.
__
] "
'
-r-
~
\ - -
"* T
:LINO
PYRO
XENE
S, LA
KE E
LLEN
KIM
BERL
ITE
Ae
clo
gite
Q
ga
rne
t py
roxe
A
sanid
ine e
clo
gite
<^s
pine
-l py
roxe
V
kya
nite
e
clo
gite
Q
peridotite
Q
meg
acry
st
1 ga
rnef
r-ep
x co
mpo
site
c
px-e
px
com
posi
te
;nite
mite
^Jxfield
of
xenocr
ysts
' 1 '
"*"
1
.......
- -
j
r "
----
-
'
*j
tn
" t
' ~
1 ' '
j_
. .
. }.
.
. ..
...
;. .
. _
4.
. . _
..- -
| t^
...j:....
..,...{
. ..
.
<~
7
-[
-
{
....
. i
....
..
. .
. .
T j_ l i
i |
567
(x ; ^ r
* '
"
~T ty\
1
3
....
..
.-
__
4-
-
' ~
-
*
;;. ;
. , .
....
.
9in w
eigh
t pe
rcen
t
Figure 7
. Wt
. %
Cr2®3
vs w
t. %
Na20 o
f cl
inop
yrox
enes
. Field
for
clin
opyr
oxen
es from
Mont
ana
garn
et pe
rido
tite
s fr
om H
earn and
McGe
e, 19
83.
>Sp
.P
rox,
> Sp
. P
er id
,
Fig
ure
8.
Thi
n se
ctio
n
phot
os o
f xe
no
lith
s fr
om
the
Lake
E
llen kim
berlite.
Ga -
ga
rne
t,
Py =
py
roxe
ne,
Ky =
ky
an
ite,
Rt
= ru
tile
, Sp
=
spin
el.
A)
Ecl
og
ite
(3
cm);
B)
Kya
nite
E
clo
gite
(1
.3
cm)
rimm
ed
by ki
mberlite
matr
ix
on
low
er
edge
;C)
G
arne
t P
yro
xen
ite
(1
cm)
en
ite -
P
erid
otite
(1
.3
cm)
serp
enitn
e in
lo
wer
right
po
rtio
n;
D)
Spi
nel
Pyr
ox-
serp
entin
e
in
uppe
r p
ort
ion
.
1400
1300
.
1200
.
1100
A E
clogite
(E
G)
^K
ya
nite
E
clogite
(E
G)
<JS
am'd
ine
Ecl
ogite
(E
G)
DD
Garn
et
Pyr
oxe
nlte
(E
G)
QS
pin
el
Pyro
xenite-P
erldotite
(W
)O
d in
opyr
oxen
e M
egac
ryst
s (L
O)
OC
1 In
opyr
oxen
e in
P
eridotite
(l_
D)
6
Ga rn
et-
Cl i
nopy
roxe
ne
(pe
r1d
ot1
te)(
LD
) -O
- C
linop
yrox
ene-
Ort
hopy
roxe
ne
(peridotite
)(W
)
S
1000 900
' ' M
onta
na
hig
h
T-P
>f
\ peri d
oti
tes
lith
lc
incl
usi
on
s co
mpo
site
xe
nocr
ysts
O
rtho
pyro
xene
X
enoc
ryst
s (B
N-M
)
Col
o.-W
yo.
garn
et
pe
rld
otite
s
Mon
tana
lo
w T
-P
X /
p
erld
otite
s
Col
o.-W
yo.
ga
rne
t p
yro
xen
ltes
Bund
y et
al,
1961
Ken
nedy
A
Ken
nedy
, 19
7fr
Pre
ssur
e (k
b)
Fig
ure
9.
Tem
pera
ture
-Pre
ssur
e D
iagr
am.
The
rmob
arom
eter
key
s:
EG=
Ell
1s «
G
reen
, 19
79;
LD-
Lin
dsl
ey
A D
1xon
19
76;
M«
Mac
Gre
gor,
19
74;
W=
We
lls,
1977
; BN
= Bo
yd &
N
ixon
, 19
73.
Dia
mon
d-gr
aphi
te b
ound
arie
s fr
om B
undy
an
d o
the
rs
(196
1)
and
Ken
nedy
A K
enne
dy
(19
76
).
P-T
fie
lds fo
r C
olor
ado-
Wyo
min
g garn
et
pyr
oxe
nlte
s an
d garn
et
pe
rld
otite
s
from
Eggle
r,
McC
allu
m,
and
Sm
ith
(19
79
).
P-T
fie
lds fo
r M
onta
na
garn
et
pe
rld
otite
s
from
Hea
rn a
nd M
cGee
(1
98
3).
Tab
le
1.M
egac
ryst
s fr
om
the
Lake
E
llen
K
imberlite
Tab
le
2A.
Com
posi
te
Xenocr
ysts
: G
arn
et-
Clin
opyr
oxe
ne
Tab
le
2B.
Com
posi
te
xenocr
ysts
: C
linop
yrox
ene-
Ort
hopy
roxe
ne
Si0
2A1
203
FeO
MgO CaO
Na20
Ti02
MnO
Cr2
03Ni
O
Si Al Fe Mg Ca Na Ti Mn Cr Ni (FeO
)(F
e^-j
GAR
H80
-2A
-2(3
)
41.4
220
.99
8.23
19.2
55.
81 -_ .85
.50
3.80 __
1M78
5
[12]
2.95
91.
767
.491
2.05
0.4
44 .045
.030
.214 ..
S70U
O
)
PYX
H80
-6A
B-1
(3)
54.4
51.
222.
9017
.30
21.3
6.8
9.2
2.1
1.1
9__
9875
4"
[6]
1.99
6.0
53.0
89.9
45.8
39.0
63.0
06.0
03.0
06 47
000"
ILM
H80
-9(3
) .09
.51
27.5
714
.62
.03
-.54
.37
.37
1.13 .1
798
759
[3]
.002
.014
.534
.505
.001 __ .947
.007
.021
.003
7703
T
22.3
65.
79
ILM
H80
-2A
I-1
(3) .05
.39
28.8
113
.94
.02
__55
.22
.36
.60
.10
9973
9
[3]
.001
.011
.557
.480
.001 __ .959
.007
.011
.002
7707
7
24.3
84.
92
GAR=
ga
rne
t,
PYX=
py
roxe
ne,
ILM
= ilm
enite
()
= n
umbe
r o
f p
oin
ts
per
gra
in.
[]=
nu
mbe
r o
f ox
ygen
s us
ed fo
r ca
tion
cal
cu
latio
n.
(FeO
) an
d (F
e20
3)
are
calc
ula
ted
va
lues.
--
not
anal
yzed
* A
ll an
alys
es
in
tab
les
1-6
wer
e do
ne
usin
g an
AR
L-
EMX
or
ARL-
SEM
Q e
lect
ron
m
icro
prob
e opera
ting
at
15
kv w
ith
a 1
mic
ron
beam
and
w
ith on-lin
e d
ata
redu
c
tion
usin
g th
e
Ben
ce-A
lbee
(1
968)
m
etho
d.
Sta
ndar
ds
used
wer
e a
ll n
atu
ral
or
syn
the
tic
min
era
ls w
ith
the
exc
ep
tion
of
a ba
rium
silic
a
glas
s st
anda
rd
(use
d fo
r B
a).
In
tab
les
1 &
2 a
ll th
e an
alys
es
are
ave
r
ages
of
seve
ral
po
ints
on
a
gra
in.
The
anal
yses
in
ta
ble
s 3-
6 ar
e av
erag
es of
seve
ral
gra
ins
in
a sa
m
ple
, w
ith
3 p
oin
ts
anal
yzed
on
ea
ch
gra
in.
Gar
nets
Si0
2A1
203
FeO
MgO
CaO
Ti0
2M
nO Cr2
03
Si Al Fe Mg
Ca Ti Mn
Cr
EG
TG-2
(3)
40.3
821
.80
18.1
614
.51
3.46
0.11
0.41
0.00
98.8
3
3.02
21.
922
1.13
61.
619
0.27
70.
006
0.02
50.
000
8.00
7
PG
TG-1
(3)
40.9
320
.52
12.6
716
.43
5.97
0.08
0.40
1.57
9O
T
3.03
21.
791
0.78
51.
814
0.47
30.
004
0.02
40.
091
8.01
4
LG
TG-4
(3)
41.7
920
.65
6.81
20.0
66.
150.
420.
242.
5698
.68
3.02
01.
757
0.41
12.
159
0.47
50.
023
0.01
50.
146
8700
5"
LG
P5-3
(3)
41.3
319
.43
6.30
20.5
14.
610.
030.
715.
4898
.40
3.00
71.
666
0.38
32.
224
0.35
90.
001
0.04
30.
315
7795
8"
LG
P9-2
(3)
41.5
020
.40
6.37
20.8
44.
220.
001.
084.
0098
.41
3.00
61.
742
0.38
62.
250
0.32
70.
000
0.06
60.
228
8700
5"
LG
S-1A
(3)
42.4
519
.83
8.14
19.9
74.
530.
100.
464.
9510
0741
"
3.03
51.
671
0.48
62.
129
0.34
70.
005
0.02
80.
279
7795
0"
Si0
2A1
203
FeO
MgO
CaO
Na20
T102
MnO
Cr2
03
Si Al Fe Mg Ca Na Ti Mn Cr
1BP-
1(3
)
52.0
84.
601.
7015
.74
22.0
5.9
9.0
7.1
11.
3598
759
1.91
0.1
99.0
52.8
60.8
67.0
71.0
02.0
03.0
39
1BP-
2(3
)
52.2
24.
411.
5815
.76
22.4
21.
00 .06
.11
1.11
9876
7
1.91
6.1
91.0
48.8
62.8
81.0
71.0
02.0
03.0
32
Cl i
nopy
roxe
nes
1BP-
3(3
)
51.8
14.
601.
6515
.60
22.1
7.9
6.1
0.0
91.
169S
7T4-
1.91
1.2
00.0
51.8
58.8
76.0
68.0
03.0
03.0
34
1A-2
2(3
)
52.5
54.
691.
6015
.49
21.6
71.
08 .10
.10
1.30
98.5
8
1.92
4.2
02.0
49.8
46.8
50.0
77.0
03.0
03.0
38
4A-1
3(3
)
52.2
84.
721.
6815
.66
21.8
0.9
4.0
9.1
01.
2298
74"9
1.91
8.2
04.0
51.8
56.8
57.0
67.0
02.0
03.0
35
1A-1
1(3
)
52.6
44.
341.
4515
.94
22.0
4.9
4.0
8.1
01.
1998
.72
1.92
5.1
87.0
44.8
69.8
63.0
66.0
02.0
03.0
34
81-4
F(3
)
52.7
24.
131.
6516
.02
21.9
1.9
2.0
6.1
2.9
698
.49
1.93
2.1
78.0
51.8
75.8
60.0
66.0
02.0
04.0
28
Clin
op
yro
xen
es
TTQo
T T7
555"
TT
QW
3.99
2
Ort
hopy
roxe
nes
1799
3"
1795
6"
S102
A
l,0
3Fe
OM
gO CaO
Na20
Ti0
2M
nOC
r20
3
Si Al Fe Mg Ca Na Ti Mn
Cr
PY6-
2(3
)
55.1
1 5.
266.
5612
.34
16.1
64.
20 .30
.02
.10
1007
05
2.00
1.2
25.1
99.6
65.6
27.2
95.0
08.0
01.0
034.
024
PY6-
1(3
)
55.3
0 .9
23.
4616
96
22.6
8.7
3.1
3.0
8.3
710
0763
1.99
7.0
39.1
04.9
13.8
77.0
50.0
03.0
02.0
1037
995-
PY6-
4(3
)
55.9
8 3.
263.
1816
.58
17.7
72.
11 .26
.00
1.23
1007
37
1.99
8.1
37.0
94.8
82.6
79.1
45.0
06.0
00.0
343.
975
GP5
-3(3
)
53.9
6 3.
601.
8714
.00
16.7
43
84 .15
.29
4.45
9879
0
1.97
0.1
54.0
56.7
62.6
54.2
72.0
04.0
08.1
284.
008
GP9
-2(3
)
55.0
8 4.
782.
8413
.78
16.0
04.
40 .08
.36
4.08
10T7
?0
1.97
3.2
02.0
60.7
35.6
14.3
05.0
02.0
05.1
15-4
70TT
GS-
1A(3
)
54.0
1 6.
102.
4712
.36
15.6
54.
83 .24
.11
3.87
9975
4
1.95
4.2
60.0
74.6
66.6
06.3
38.0
06.0
03.1
10?7
0T7
Si 0
2A1
203
FeO
MgO
CaO
Na20
Ti0
2M
nO Cr2
03
Si Al Fe Mg Ca Na Ti Mn Cr
(6)
54.3
84.
48
5.37
33.0
8.6
0.0
4.0
4.1
4.7
898
.91
1.89
4.1
84.1
561.
718
.022
.003
.001
.004
.022
T70U
4"
(6)
54.2
64.
51
5.25
32.7
1.8
0.0
4.0
4.1
4.8
398
758
1.89
6.1
86.1
531.
704
.030
.003
.001
.004
.023
4TD
M
(2)
54.1
74.
16-
5.44
33.0
2.3
8.0
2.0
0.1
5.7
998
.13"
1.90
1.1
72.1
601.
728
-014
.001
.000
.004
.022
T70
0?
(3)
54.8
94.
10
5.30
33.4
8.3
2.0
2.0
2.1
4.7
298
.99
1.90
7.1
68.1
541.
733
.012
.002
.001
.004
.020
4700
1
(3)
55.0
54.
60
5.52
33.1
2.4
1.0
2.0
1.1
3.6
999
755
1.90
2.1
88.1
601.
706
.015
.001
.000
.004
.019
1799
5"
(3)
54.8
44.
12
5.35
33.4
8.2
6.0
1.0
2.1
3.6
598
756
1.90
7.1
69.1
561.
735
.010
.001
.000
.004
.018
4.00
0
(6)
54.8
04.
41
5.40
33.3
8.2
9.0
3.0
5.1
3.7
499
72"3
1.90
0.1
80.1
571.
725
.011
.002
.001
.004
.020
T70
W
E= e
clo
gite
, P=
garn
et
pyr
oxe
m'te
, L-
()
=
num
ber
of
poin
ts
per
gra
in.
garn
et
pe
rid
otite
()=
nu
mbe
r of
poin
ts
per
gra
in.
TABL
E 3.
G
arne
ts
from
Lithic
In
clu
sio
ns,
La
ke E
llen
Kim
berlite
Tab
le 4
. P
yrox
enes
fr
om L
ith
ic
Inclu
sio
ns,
Lake
Elle
n K
imberlite
Si 0
2 A1
203
FeO
M
gO
CaO
Na20
MnO
C
r203
KB
2-1
E (8)
40.0
8 22
.72
17.7
8 10
.10
9.15 .0
9 .2
9 .1
1 H
JOT3
?
KB
2-2
E (7)
40.6
9 22
.75
15.4
9 12
.70
7.88 .0
9 .2
0 .0
9
KB
2-3
K (7)
40.5
1 22
.48
15.1
4 12
.69
8.13 .1
0 .2
1 .0
6 55
73?
H80
- 1A
-2
P (5)
41.4
5 23
.11
13.9
2 15
.51
5.71 .0
5 .3
4 .4
4 10
0.53
H80
- 1A
-1
E (9)
40.5
0 22
.81
16.6
1 10
.02
10.4
5
.13
.30
.06
100.
88
H80
- 2A
-1
E (6)
39.7
3 22
.78
18.6
9 9.
07
9.80 .1
1 .4
4 .0
8 1T
JOT7
D"
H80
- 1A
-4
E (4)
40.5
8 23
.15
16.2
2 14
.41
5.28 .0
9 .3
2 .1
3 10
0.18
H80
- 6A
-3
S (3)
40.7
5 23
.58
15.9
0 12
.60
8.23
.0
6.0
9 .2
2 .0
4 1U
T747
H81
- 4D
-1
S (2)
39.6
4 22
.81
19.5
8 8.
80
9.73
.0
7.1
5 .4
1 .0
4 10
1.23
H81
- 4A
-1
S (4)
39.3
4 22
.38
20.8
3 8.
13
8.82
.0
7.1
0 .4
3 .0
1 10
0.11
Si 0
2 A1
203
FeO
M
gO
CaO
Na?
0 T
i02
MnO
KB
2-1
E (10)
54.2
1 13
.93
4.00
8.
33
13.4
46.
22
.33
.07
.04
10U
757
KB
2-2
E (17)
55.7
5 13
.39
2.52
8.
69
13.0
76.
96
.19
.05
.13
100.
75
KB
2-3
K (16)
54.8
6 14
.65
2.39
8.
29
12.7
86.
88
.30
.08
.05
100.
28
H80
- 1A
-2 P (8)
53.5
9 3.
97
2.90
15
.77
22.8
1.8
5 .3
7 .0
6 .5
3 10
0.85
H80
- 1A
-1 E (8)
55.2
1 14
.30
3.11
7.
23
12.6
17.
08
.28
.08
.03
99.9
3
H80
- 1A
-4 E (5)
55.4
6 9.
97
3.54
10
.13
14.2
06.
08
.26
.07
.13
99.8
4
H80
- 2A
-1 E (8)
56.0
1 15
.07
3.24
6.
59
11.1
28.
11
.27
.11
.13
100.
65
H80
- 6A
-3 S (6)
55.7
9 13
.99
2.44
7.
11
12.0
57.
42
.18
.07
.02
99.0
7
H81
- 4D
-1 S (2)
56.6
1 14
.56
3.84
5.
87
11.1
37.
62
.24
.08
.07
100.
02
H81
- 4A
-1 S (2)
55.8
4 12
.62
4.82
6.
42
12.1
57.
24
.20
.11
.12
H80
- 4B
-1 B (10)
52.6
4 4.
38
3.32
15
.59
22.6
91.
02
.34
.11
.38
1W74
T
H80
- 4B
-1 B (4)
54.7
5 3.
14
8.96
31
.72
.25
.00
.07
.22
.16
99.2
7
Si
Al
Fe
Mg Ca
Na
Ti
Mn Cr
E=
eclo
gite
, K=
ky
an
ite eclo
gite,
P=
garn
et
pyr
oxe
nite
, S=
sa
nid
ine eclo
gite
Ca
tion
s ar
e ba
sed
on
12
oxyg
ens.
Sam
ple
H80
-6A
-3 c
on
tain
s bo
th ky
an
ite
and
san
idin
e.
()=
nu
mbe
r of
gra
ins
per
sam
ple.
--
no
t an
alyz
ed.
£.»»
/ 2.
000
1.11
0 1.
125
.732
.005
.0
18
.007
7755
4
J.U
iU
1.98
1 .9
56
1.39
9 .6
23
.005
.0
12
.005
7755
T
J.U
IO
1.96
8 .9
40
1.40
6 .6
46
.006
.0
13
.004
7755
6"
o.uu
/ 1.
974
.843
1.
676
.443
.003
.0
21
.025
7.99
2
J.U
UH
1.99
2 1.
029
1.10
7 .8
29
.007
.0
19
.003
7755
0"
C..IQ
C.
2.01
2 1.
171
1.01
4 .7
86
.006
.0
28
.005
8.00
4
c.yo
-3
2.01
9 .9
95
1.57
8 .4
15
.005
.0
20
.008
8.02
3
£.3
/H
2.02
6 96
9 1.
370
.642
.0
09.0
05
.014
.0
0257
0IT
C..-3
I I
2.01
2 1.
225
.982
.7
80
.010
.008
.0
26
.004
570T
8"
£.3
3J
2.00
5 1.
323
.921
.7
17
.010
.006
.0
28
.006
O0
9~
Si
Al Fe
Mg Ca Na
Ti
Mn
Cr
1 91
6 .5
79
.118
.4
39
.508
.425
.0
09
.002
.001
37
557
1.95
7 .5
52
.074
.4
53
.489
.471
.0
05
.002
.004
4.
007
1.92
7 .6
06
.070
.4
34
.480
.467
.0
08
.003
.001
17
556"
1.93
1 .1
68
.087
.8
46
.879
.059
.0
10
.022
.015
T
70T
7
1.95
0 .5
94
.092
.3
80
.476
.484
.0
07
.002
.001
37
556"
1.97
8 .4
18
.105
.5
38
.542
.419
.0
07
.002
.004
T
70T
T
1.95
9 .6
20
.094
.3
43
416
.549
.0
07
.003
.004
3.
995
1.97
7 .5
84
.072
.3
75
.456
.509
.0
05
.002
.001
3.
981
1.99
0 .6
03
.113
.3
07
.418
.518
.0
06
.002
.002
3.
959
1 99
3 .5
30
.144
.3
41
.464
.500
.0
05
.003
.003
37
553"
1.91
1 .1
87
.100
.8
43
.880
.072
.0
09
.003
.011
T7
0T5"
1.92
7 .1
30
.263
1.
662
.009
.000
.0
02
.006
.004
47
0-03
"
E= e
clo
gite
, K=
ky
an
ite e
clo
gite
, P=
g
arn
et
pyr
oxe
nite
, S=
sa
nid
ine eclo
gite,
B=
spin
el
pyr
oxe
nite
Ca
tion
s ar
e ba
sed
on 6
ox
ygen
s.S
ampl
e H
80-6
A-3
co
nta
ins
both
ky
an
ite
and
sanid
ine.
Sam
ple
H80
-4B
-1
anal
yses
ar
e fo
r cl
ino
pyr
oxe
ne
and
ort
hopyr
oxe
ne.
()=
nu
mbe
r o
f gra
ins
anal
yzed
pe
r sa
mpl
e.
Tab
le
5.
Oxi
des
and
kyam
'te
from
lith
ic in
clu
sio
ns,
Lake
E
llen
K
imberlite
Tabl
e 6.
Feld
spar
s fr
om ecl
ogit
e inclusions,
Lake El
len
Kimberl He.
Si0
2A1
203
FeO
MgO CaO
Ti0
2M
nOC
r203
NiO
Si Al Fe Mg Ca Ti Mn Cr Ni (FeO
)(F
e20
3)
RUT
K82
-1 E (1) .. .27
.82
.00
99.9
8.0
0.0
8.0
0
__
.004
.009
.000 .991
.000
.000
.000
T70
04
RUT
KB
2-2
E (1) __ .17
.26
.00
98.1
2.0
0.3
2.1
399
.00
.. .002
.002
.000 .993
.000
.003
.001
T70
0T
RUT
KB
2-3
K (2) ._ .30
.32
.00
98.8
0.0
0.1
2 1
899
.72
__ .005
.004
.000 .993
.000
.001
.002
1.00
5
RUT
H80
-1A
-1E (6) .84
.33
1.14 .06
.05
96.2
4.0
6.1
1.0
198
754"
.011
.005
.013
.001
.001
.976
.001
.001
.000
T70
05
RUT
H80
-1A
-4E (2) .98
.14
.58
.06
.42
97.0
4.0
5.2
5.0
499
756"
.013
.002
.006
.001
.006
.976
.001
.003
.000
T70
08
RUT
H80
-2A
-1 E (4) .97
.28
.51
.04
.20
97.2
6.0
7.2
0.0
199
.54
.013
.004
.006
.001
.003
.977
.001
.002
.000
T700
7
SPN
H80
-1A
-2P (3) .00
55.6
916
.69
17.5
5.1
1.0
4.0
09.
72 .46
ITJO
TZB"
.000
1.74
3.3
70.6
95.0
03.0
01.0
00.2
04.0
10T7
0Z6
13.5
93.
44
SPN
H80
-4B
-1B (7) .00
55.3
418
.92
16.6
7-- .0
5-- 8.76 .3
51T
507U
5
.000
1.74
8.4
24.6
66 .001 .185
.008
1703
2
15.0
94.
25
COR
KB2
-3K (2) _
98.3
9.3
5.1
8-- .2
5.0
0.2
4.0
855
7*5
__1.
985
.005
.005 .003 .003
.001
1700
2
KY
KB
2-3
K (9)
37.5
662
.89
.34
.08
.08
_- _- .09
__10
1.04
1.00
51.
982
.008
.003
.002 -- .002 _-
3700
7
RUT= ru
tile
(2); SPN= s
pinel
(4);
COR= c
orun
dum
(3); KY
= kyanite
(5);
number
indi
cate
sth
e nu
mber
of
oxygens
used
to
calculate
cati
ons.
(FeO)
and
(Fe203)
are
calc
ulat
ed values.
E= e
clog
ite,
K=
kyanite
eclo
gite
, P= g
arne
t pyroxenite,
B= sp
inel
pyroxenite
()=
numb
er o
f gr
ains
per
sample.
--
not
anal
yzed
.
SiO
?A1
203
FeO
MgO
CaO
Na^
K20
BaO
Si Al Fe Mg Ca Na K Ba Or Ab An
H80
-6A
-3(2
)
62.7
019
.21
.00
.00
.02
.30
15.4
81.
9899
759
2.94
61.
063
.000
.000
.001
.027
.926
.036
'475
55
97.1 2.8
0.1
H80
-6A
-4(4
)
64.1
418
.72
.01
.00
.00
.39
16.0
1.3
85O
B"
2.98
21.
025
.000
.000
.000
.035
.948
.007
"475
57
96.4 3.6
0.0
H81
-4A
-1(1
)
64.9
518
.75
.01
.00
.02
1.30
14.4
7.4
159
75T
2.99
11.
018
.000
.000
.001
.116
.850
.007
"475
13
87.9
12.0
0.1
H81
-4D
-1(3
)
63.9
118
.68
.00
.00
.00
.34
16.0
7.3
7S5
737
2.98
11.
026
.000
.000
.000
.031
.954
.007
4.99
9
96.8 3.1
0.0
H81
-4A
-6-
2-2
64.1
318
.83
.00
.00
.00
.79
14.9
3.5
799
.25
2.98
21.
032
.000
.000
.000
.071
.885
.010
"479
80
92.6 7.4
0.0
2-3
65.7
119
.00
.02
.00
.00
2.40
12.8
3.6
610
0762
2.99
01.
019
.001
.000
.000
.212
.745
.012
"475
75
77.8
"22.
20.
0
Cations
are
based
on 8 ox
ygen
s.H81-4A-6 s
howe
d some zoning in t
he feldspar,
sing
leanalyses o
f th
e tw
o extremes w
ithi
n on
e grain
are
shown
above.
()=
numb
er o
f gr
ains
per
sample.