Marine Geology, 22 (1976) 271--283 © Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands
AN E L E C T R O N MICROPROBE STUDY O F G L A U CO N IT E S FROM T H E C O N T I N E N T A L MARGIN O F F T H E WEST COAST OF SOUTH A F R I C A
G.F. BIRCH', J.P. WILLIS2 and R.S. RICKARD 2 ' Marine Geoscience, Department of Geology, The University of Cape Town, Cape Town (South Africa) 2Department of Geochemistry, The University of Cape Town, Cape Town (South Africa) (Received April 13, 1976; revised and accepted July 26, 1976)
ABSTRACT Birch, G.F., Willis, J.P. and Rickard, R.S., 1976. An electron microprobe study of glauconites from the continental margin off the west coast of South Africa. Mar. Geol., 22: 271--283. The rich giauconite deposits off the south and west coasts of southern Africa originated by erosion of Cretaceous and Tertiary bedrock during mainly Tertiary transgressive/ regressive cycles. This work, the first major electron microprobe investigation of glauconite, was undertaken to provide reliable chemical data of uncontaminated, unaltered glauconite pellets and to trace the chemistry of the glauconitization process. The data suggest that Fe is emplaced into the clay structure very early in the glauconitization process, possibly by a mechanism which is independent of and prior to the fixation of K. The sympathetic relationship between K:O and MgO indicates that some Mg is located in the interlayer position. INTRODUC~ON
Rich giauconite deposits cover extensive areas of the continental margin o f f the south and west coasts o f southern Africa (Birch, 1971, 1975; Rogers, 1974; Bremner, 1975). The glauconite is c o n c e n t r a t e d in belts parallel to the coast. In these areas glauconite constitutes over 50% of the total sediment and values o f greater than 90% are c o m m o n (Fig.l). The glauconite has been eroded f r o m Cretaceous and Tertiary sediments during repeated transgressions and regressions, mainly during t he Tertiary. During these periods t he less dense constituents of the bedr oc k were disaggregated and the finer material winnowed seawards, leaving the glauconite c o n c e n t r a t e d in hollows in the b e d r o c k as a lag-placer deposit. L ow terrigenous and pelagic sedimentation (Birch, 1975) has reduced dilution so t h a t the deposits have been preserved. An anomalously deep shelf (Rogers, 1974; Bremner, 1975) and Late Tertiary downwarping in t he south west (Dingle, 1971) has lowered m ost of t he glauconite deposits t o depths b e y o n d the influence of Pleistocene sea-level fluctuations. Only the glauconites o f f Port Elizabeth, on the south coast, were affected by post-Tertiary sea-level changes.
272
15° I
20 ° I
¢.
25 ° I
30° I
35 ° I
,20 o
I ws,v~.say
over 50% of total sediment over 25% of total sediment area studied samples selected microprobe
LUderitz_-_
i L~j~ .25 °
for
analysis
Lourenco Marquea~ /
shelf break
/
/
.30 °
-35 °
I
I
I
Fig.1. Glauconite deposits off the coasts of southern Africa.
The term "glauconite" has previously been employed as a name for specific mineral species and as a morphological term for greenish, round to lobate pellets most commonly of sand size (Burst, 1958a,b). In this paper the term "glauconite" is applied in the morphological sense and the term "mineral gtauconite" is reserved for the monomineralic micaceous clay comprising only non-expandable illite-like minerals and containing more than 8% K20. Obtaining reliable chemical data on glauconite is an extremely difficult task (Burst, 1958b; Bentor and Kastner, 1965; Birch and Willis, 1974) not only because every contaminated, oxidized and altered grain must be extracted from a bulk sample by hand, b u t also because many grains are composite and contain impurities in the form of other mineral phases. It is often impo~ible to hand pick samptes as impurities are sometimes only discernible in thin-section or by X-ray analysis. In spite of these difficulties,
273 previous geochemical studies made by Birch (1971) showed that South African glauconites manifested chemical and mineralogical trends controlled by diagenetic processes. The main purpose of the present study was to acquire reliable chemical data on uncontaminated, unaltered pellets and to trace the chemistry of the glauconitization process using electron microprobe techniques. The formation, diagenetic development and age of the glauconitic material discussed here is beyond the scope of this paper and is presented elsewhere (Birch, 1975, Birch, in prep.). Only two other minor microprobe investigations (5 determinations) have been made on glauconitic material. A single analysis of a "glauconie" foraminiferal infilling was reported by Bjerkli and Ostmo-Saeter (1973), and Burnett (1974) has presented four analyses of glauconite from the Peruvian--Chilean shelf. Mineralogical data can be taken from Birch (1971) because to acquire structural data compatible with the microprobe determinations would necessitate using special techniques involving single grain analyses which were not available to us. METHODS Glauconite grains were mounted in araldite, thin-sectioned and doublepolished before mounting on graduated slides for easy grain identification. Minerals were analysed for Si, A1, Fe, Mg, Ca, Na, K and P using a Cambridge Microscan V electron microprobe in the Department of Geochemistry, University of Cape Town. The raw data was corrected and reduced using ABFAN, a computer programme written by Boyd et al. (1968). Ka lines for the 8 elements were measured at 15 kV with a total beam current of 0.5 • 10 -7 A using flow proportional detectors. The following analyzing crystals were used: RAP crystal for Si, A1, Mg and Na; QUARTZ for Fe, Ca and K; and PET for P. To prevent the volatilisation of Na and K from the sample area under the electron beam, it was necessary to use a defocussed beam. Experiments showed that no loss of intensity due to defocussing the spot occurred as long as the spot diameter was less than 30 pin. Spot diameters between 10 and 20 pm were used in this work, and no measurable loss of alkali elements was noticed in the counting times employed (20 sec). MICROPROBE ANALYSES As the composition of glauconite is variable, and because no comparative data exists, it was first necessary to carry out repeated analyses of one glauconite type from a few locations to establish the degree of intergrain homogeneity and to determine the precision of the analytical method. Replicate analyses were made on 10 mature glauconite pellets from the two main glauconite deposits, and on a further 10 mature glauconite pellets from
274
other locations on the shelf. A further 27 pellets at different stages of maturity, were analysed in an effort to elucidate the chemical changes which take place on glauconitization. The glauconite grains were grouped into four stages of maturity (Fig.2) based on the following petrographic criteria (listed in order of increasing maturity): 1st stage: Light yellow- or brownish-green, slightly sutured with an irregular outline, and with common inclusions of silt-sized quartz or shell fragments.
Stage 1
Stage ~
Stage 3
Stage 4
MATURITY
100--1 -4 -4 -4
S i O 2 e-
•
-4 4
FezO3 • . . . . . . . . .
- . . . . . . . . . . . . . .
AI
K,o'"
I,-
MgO
CaO
4
: •
•
"
~
@ "'° T""
~
..0 .....
t.0o0..°.~o°o,.o
....
........
...................
0°°.°
o
°'°o, o °°°..% o,s-.t
•
°°'.Oo°. °'.Ooo °°°.. o "o ......................
.0..° "'"°'°°*°"o
(I,1-
Fig.2. The composition of glauconite groups arranged according to maturity.
•
275
2nd stage: Yellow-green, sutured, with a subrounded outline, and with no inclusions. 3rd stage: Uniform gray-green, very sutured, with rounded outline. 4th stage: Uniform dark green, non-sutured to slightly sutured, and generally well rounded. The mineral glauconite The maximum theoretical K20 content for glauconites has been claculated at 8.7%, but for A1 2:1 contracted layer minerals it can be as high as 9.9% K20 (Cimb~linkSv~i, 1971, in: Weaver and Pollard, 1973). Generally, glauconite containing more than 8% K20 or even possibly greater than 7% K20 is considered to be pure "mature glauconite" (Burst, 1958b; McRae, 1972). The most mature glauconites (4th stage) from off the Cape west coast all contain over 8% K20 (Table I), have a maximum content of 9.4% K20 and a mean value of 9.01%, so can be considered pure. Compared with the literature, this K20 content seems high, probably because the reported analyses are bulk determinations which will include some non-glauconitic material. Bulk glauconite analyses are generally contaminated by non-structural Fe or are affected by the inclusion of oxidised grains. It is therefore interesting to note the large range in Fe203 values (16.6--23.0% Fe203) that occurs in this group of analyses, and also that the average bulk determinations for the Agulhas Bank glauconites are similar to the more recent probe data, i.e., 21.22 TABLE I Average composition of glauconitic material as determined by microprobe analysis Glauconite pellets
No. of analyses
ForaAltered miniferal faecal oxidized infillings pellets
1st stage
2nd
3rd
4th
stage
stage
stage
6
12
9
20
5
3
1
48.03 5.74 19.14 4.04 1.68 0.14 5.22 0.19
49.39 4.30 22.43 4.25 0.32 0.19 6.73 0.20
50.81 4.56 21.86 4.46 0.32 0.08 7.96 0.17
50.97 4.29 21.01 5.45 0.18 0.07 9.01 0.15
28.08 4.00 44.73 2.83 0.30 0.11 3.40 0.52
47.76 6.42 19.53 4.72 2.15 0.06 6.92 0.39
54.15 5.92 19.15 4.91 0.48 0.12 8.56 0.25
84.18
87.81
90.22
91.13
83.97
87.95
93.54
Components SiO2 A1203 Fe203 MgO CaO Na~O K20
P~Os Total*
* Low totals are due to the water content which varies from up to 16% for immature glauconites to 8% for the most mature types (Birch, 1971).
276
and 21.01% Fe203 respectively. Apparently non-structural Fe and oxidized grains are not important in glauconites from this region. The variation in the SiO2 content (47.7--52.2%) is not as much as that reported by most workers (Hutton and Steelye, 1941; Bentor and Kastner, 1965), but the A1203 content ranges markedly (1.77%--7.01%). Such a variation is common because A1 can substitute for Fe in the glauconite structure (McRae, 1972; Weaver and Pollard, 1973). The low Na20 (0.07%), CaO (0.18%) and P2Os (0.15%) content indicate that the mineral is essentially free of apatite. Glauconites from the west coast of South Africa have a greater range in MgO values (3.75--6.23%) than is normal for glauconites and the average value (5.45%) is high in relation to glauconite reported in the literature (4%; McRae, 1972).
Immature glauconite and the glauconitization process Glauconites containing <8% K20 have been grouped into three classes according to maturity. The range in elemental abundance for all the glauconite analyses appears below, and the mean values for some of the elements in each group are presented in the form of a variation diagram (Fig.2). Maximum Minimum
SiO 2 52.22 43.22
A1203 7.39 1.97
Fe203 27.27 16.62
K~O 9.47 4.64
MgO 6.23 3.75
Na20 0.57 0.02
CaO 3.51 0.04
P20~ 0.58 0.03
Fig.2 and Table I show the following trends: (1) MgO increases with increasing K20. (2) SiO2 increases marginally with increasing K20. (3) CaO and AI2 03 decrease with increasing K20. (4) Fe203 displays no consistent trend with increasing K20. These trends are confirmed in the scatter diagrams of individual analyses (Figs.3 and 4). McRae (1972) has combined the various glauconitization theories (excluding precipitation) into two groups: the "layer lattice theory" (Burst, 1958a,b; Hower, 1961) and the "epigenetic substitution theory" (Ehlmann et al., 1963; Seed, 1965; and others). According to Burst's model, the formation of glauconite simply requires any degraded 2:1 layer lattice structure with a low lattice charge as a parent material. The glauconitization process involves a gradual substitution of Fe for A1 in the octahedral position, resulting in an increase in octahedral lattice charge and a consequent increase in "charge-balancing" interlayer cations (mainlyK). This causes the collapse of increasingly more expandable layers to a 10/~ non-expandable type. The alternative theory was initially proposed to explain the preferential replacement of calcite by glauconite (McRae, 1972), but has been used to explain the replacement of many non-micaceous silicates, e,g. quartz, feldspars, pyroxenes, olivines, etc. (Ojakangas and Keller, 1964; Dapples, 1967). Although there is ample evidence for both theories, there is general agreement that most glauconites are formed by alteration of micaceous
277
/ // //
% % %% •
•
•
,. ,,.7~ ." %% (~
/
•
• o
Oe
¢/
•
/ ¢
\
0 o •
•
•
i/
//
/
\
/ \
//
0
0
O/ •
•
•
o',
\
o\ /
I
.J.... i,
.... ! .... ! ....
r
,
t w
,e
0
o
o
I •
O00
/
o
•
~
.;",2.
o
//O
• •
/ /¢
•
0 e-
0 //
• oo oo
•
•
•
•
iI
// •
/
/
e
/
:
•
-- - o
//
//
I
•
•
•
i/
•
//
•
i
///
I--
°
.
Z
Fig.3. M i c r o p r o b e a n a l y s e s o f g l a u c o n i t e s , s t a g e s I to 4.
278
\\ ~\
o• •
~. ~.,.'.. ..0. °, \
•
oo
~o o
441
oXx o~
\
•
~0
\
I:
\
0
N ii-/I / iI
• "...,,,.'.:.. i~
'*t
i i/
•
\
/
/
l
//
•
•
1/
J
~ o//
i
•
" 0 • •
\
• o
e/I // t
•
// \
I ....
gob
•
•
/ .
,,¢
///
F
,
i
O
Fig.4. Microprobe analyses of glauconites, stages 1 to 4.
,
n c~
d c~ M=
•
. ~ ./
%
L ....
e/~ee e
oO/eP
% ,
~-r-T r r I-T r T ~
~
,
I
279 material. Nevertheless, the precise mechanism involved in the diagenetic change is still obscure. There is clearly disagreement with regard to the roles played by Fe and Mg in the glauconitization process. The model proposed by Burst (1958a, b) and Hower (1961) predicts that the increase in net lattice charge is caused by substitution in the octahedral layers. The number of interlayer cations (mostly K) would increase with the net octahedral charge and therefore there should be no relationship between the abundance of interlayer cations and the tetrahedral charge. If this model is correct, K and Fe should be proportional to each other and inversely proportional to the number of expandable layers. However, several recent studies (Bentor and Kastner, 1965; Foster, 1969; Weaver and Pollard, 1973) have found no such relationship and data obtained in the present investigation seem to substantiate that no relationship exists between K and Fe. An alternative theory considers the process of glauconitization to be of "two separate unrelated processes; incorporation of Fe into the crystal structure and fixation of K in interlayer positions with the incorporation of Fe and development of a negative layer charge preceding the complete fixation of K" (Foster, 1969, p.F14). Hower (1961, p.326) considers the Fe and K to be "adsorbed" contemporaneously from the sea water onto a degraded layer, but there is no evidence in the present data to suggest that this relationship exists (Fig.3). This may be due to all the Fe being assigned to the ferric state which could be masking a subtle relationship between K20 and FeO or Fe2 03. The least potassic material determined in this study (4.6% K20) contained over 19% Fe203 while the glauconite with the highest K:O content (9.4%) contains only 18.88% Fe203. These data (Figs.3 and 4) tend to support Foster's theory. Similar relationships have been found by Ehlman et al. (1963), Seed (1965), Bentor and Kastner (1965) and Foster (1969). The most immature glauconite infillings found by Ehlmann and others contained 4.31% K20 and 22.3% Fe203, and similarly Bjerkli and C)stmo-Saeter (1973) analysed a glauconie (an immature glauconite) infiUing which contained only 2.77% K20 but 18.71% FeO. Ross and Hendriks (1945) allocated all the MgO in glauconite to the octahedral layer and obtained octahedral occupancy of as high as 2.24 positions. From this they concluded that octahedral occupancy in some dioctahedral minerals is not necessarily limited to two-thirds of the available cationic octahedral positions, but that as much as a fourth of the "vacant" third positions might actually be occupied. Bentor and Kastner (1965) also found an "excess" (between 2 and 33% of the total Mg2÷ present) of Mg, but explained it in another way. Only by assuming that some of the Mg2÷ in the octahedral layer was not replacing A13÷but was adding to it, could the negative lattice charge be balanced by the amount of adsorbed I~ cations present. The (060) position of glauconite {1.51--1.53/~) is intermediate between that of a dioctahedral (1.50 .~) and a trioctahedral (1.54/~) 2:1 layer clay. Warshaw (1957) and Bentor and Kastner (1965) present this as independent evidence for some trioctahedral layers being intergrown with octahedral layers.
280
MgO is invariably assigned completely to the octahedral layer; A13÷ and Fe 3÷ account for approximately 80% of the positions in the octahedral sheet, (Jung, 1954, in: Millot, 1970) and Mg2÷ and Fe 2÷ constitute about equal proportions of the remainder (Foster, 1951; Weaver and Pollard, 1973). K~ster {1965) plotted MgO against KEO to show that Mg in the octahedral layer was directly related to the quantity of interlayer K. Similar evidence (Foster, 1951; Owens and Minard, 1960; and others) of exchangeable interlayer MgO has led Foster (1969) to consider all Mg in excess of two octahedral cations as being present in the interlayer position. The concentration of all elements would increase with increasing K20 because of the dehydration that accompanies the glauconitization process, but the marked enrichment in MgO with increasing K20 (figs.2 and 4) as well as the unusually high MgO values determined in this investigation would favour the supposition that some Mg is being incorporated with K in the interlayer position. Ca and Na are usually assumed to be present as exchangeable cations firmly fixed in the interlayer position (Weaver and Pollard, 1973). The only relationship between the interlayer cations exhibited by these data is a general decrease of Na and Ca with increasing K (Fig.3). This suggests that K and Mg increase in the interlayer position at the expense of Ca and Na. High Ca values in some immature glauconites (i.e. up to 0.5%) are not associated with interlayer adsorption or an apatite phase (P2Os values <0.13%), but are due to the presence of minor calcium carbonate, as determined by X-ray image scanning.
Glauconitic foraminiferal in fillings and faecal pellets Analyses of infillings and of glauconitized faecal pellets (Table I) produce results similar to those obtained for the glauconite grains described earlier and serve only to confirm the presence of abundant Fe as well as that of calcium carbonate in immature glauconite. It is also interesting to compare these results with those presented by Bjerkli and Ostmo-Saeter (1973) for a foraminiferal infilling from off the Scandinavian coast (Table II). The presence of abundant Fe and high Ca (probably as CaCO3 as indicated by X-ray image scans) in the most immature glauconite phase is substantiated by both studies, and the composition of the glauconitic faecal pellet is similar to that of the fourth stage glauconites. CONCLUSIONS
Practically all chemical analyses of glauconite have been carried out on bulk samples. In this paper we report the results of the first large-scale electron microprobe study of glauconite material. Bulk determinations, however carefully conducted, are susceptible to error due to contamination by oxidized or altered pellets or even by non-glauconite minerals. In addition bulk analyses reflect the average composition of all
281 TABLE II Foraminiferal infillings from off Norwegian and South African coasts. .B.jerkli and Ostmo-Saeter (1973)
SiO2 A1203 Fe203 MgO CaO Na20 K:O
38.56 6.06 18.71" 2.92 8.57 trace 3.77
P20s
--
Glauconite infillings determined in the present study most immature
most mature
45.21 6.14 20.15 4.21 5.07 0.05 5.92
50.18 6.41 18.21 5.43 0.88 0.08 8.87
0.35
0.66
* Total Fe as FeO.
grains comprising the sample. These pellets may range in composition from immature, hydrated mixed mineral clay aggregates to mature potassium-rich pellets of mineral glauconite. The unusually high K content of the mature glauconite (9.01%) from off the Cape west coast is due to the sample being undiluted by impure material. Low P2Os and Na20 values indicate that phosphate contamination, which is a c o m m o n problem (Birch, 1971), has been practically eliminated. The chemical changes which take place during glauconitization have been monitored by careful selection of glauconite pellets for microprobe analysis. Glauconite samples where chosen at four stages of maturity based mainly on the change in colour (from light yellow-green to dark green), and morphology which the grain undergoes during the diagenetic process. The results show that the Fe is emplaced into the clay structure very early in the diagenetic process by a mechanism which is independent of the fixation of K. This evidence favours the model of glauconitization proposed by Foster (1969) rather than that described by Burst (1958a) who envisaged the introduction of K being due to a charge imbalance resulting from Fe/A1 substitution. If this were the case, Fe would be positively related to K and such an association is clearly absent. A sympathetic relationship between K20 and MgO and the anomalously high MgO values suggest that some of the Mg is located in the interlayer position with K. ACKNOWLEDGEMENTS
We thank Professor R.V. Dingle, University of Cape Town, for criticism of a draft of the manuscript and Mrs G. Krummeck for typing. Messrs. D. Wilson and R. Mitchell, U.C.T. helped in the sample preparation and Miss J.A. Chiddy undertook the drafting. The Director of the Geological Survey
282
of South Africa is gratefully thanked for allowing G.F. Birch to work on this paper. J.P. Willis wishes to thank the C.S.I.R., Pretoria and the University of Cape Town for financial support.
REFERENCES Bentor, Y.K. and Kastner, M., 1965. Notes on the mineralogy and origin of glauconite. J. Sediment. Petrol., 35(1): 155--166. Birch, G.F., 1971. The glauconite deposits on the Agulhas Bank, South Africa. SANCOR Mar. Geol. Prog. Bull. No. 4, Geol. Dep. Univ. Cape Town, 134 pp. Birch, G.F., 1975. Sediments on the Continental Margin off the West Coast of South Africa. Thesis, Univ. of Cape Town, 210 pp. (unpublished). Birch, G.F. and Willis, J.P., 1974. Electron microprobe studies of marine glauconite and phosphorite. Proc. Electron Microsc. Soc. S. Afr., 4: 79--80. Birch, G.F., in prep. Age and diagenetic development of glauconite from the west coast off South Africa. Bjerkli, K. and C)stmo-Saeter, J.S., 1973. Formation of glauconie in foraminiferal shells on the continental shelf off Norway. Mar. Geol., 14: 169--178. Boyd, F.R., Finger, L.W. and Chayes, F., 1968. Computer reduction of electron probe data. Carnegie Inst. Washington Yearb., 67: 210--215. Bremner, J.M., 1975. Faecal pellets, glauconite, phosphorite and bedrock from the Kunene-Walvis continental margin. GSO/UCT Mar. Geol. Prog. Tech. Rep., Geol. Dep. Univ. Cape Town, 7: 59--68. Burnett, W.G., 1974. Phosphorite Deposits from the Sea Floor off Peru and Chile: Radiochemical and Geochemical Investigations Concerning their Origin. Thesis, Hawaii Institute of Geophysics, Univ. of Hawaii, 164 pp. (unpublished). Burst, J.F., 1958a. Glauconite pellets. Their mineral nature and application to stratigraphic interpretations. Bull. Am. Assoc. Pet. Geol., 42: 310--327. Burst, J.F., 1958b. Mineral heterogeneity in glauconite pellets. Am. Mineral., 43: 481--497. Dapples, E.G., 1967. Diagenesis of sandstones. In: G.V. Chilingar (Editor), Diagenesis in Sediments. Elsevier, Amsterdam, pp.91--125. D:.ngte, R.V., 1971. Tertiary sedimentary history of the continental shelf off southern Cape Province, South Africa. Trans. Geol. Soc. S. Aft., 74: 173--186. Ehlmann, A.J., Hulings, N.C. and Glover, E.D., 1963. Stages of glauconite formation in modern foraminiferat sediments. J. Sediment. Petrol., 33: 87--96. Foster, M.D., 1951. The importance of exchangeable magnesium and cation-exchange capacity in the study of montmorillonitic clays. Am. Mineral., 36: 717--730. Foster, M.D., 1969. Studies of celadonite and glauconite. U.S. Geol. Surv. Prof. Pap., 614-F: 17 pp. Hower, J., 1961. Some factors concerning the nature and origin of glauconite. Am. Mineral., 46: 313--334. Hutton, C.O. and Steelye, F.T., 1941. Composition and properties of New Zealand glauconites. Am. Mineral., 26(10): 595--605. KSster, H.M., 1965. Glaukonit aus der Ragensburger Oberkreideformation. Beitr. Mineral. Petrogr., 11: 614--620. McRae, S.G., 1972. Glauconite. Earth-Sci. Rev., 8: 397--440. Millot, G., 1970. Geology of Clays. Springer, New York, N.Y., 429 pp. Ojakangas, R.W. and Keller, W.G., 1964. Glauconitization of rhyolite sand grains. J. Sediment. Petrol., 34: 84--90. Owens, J.P. and Minard, J.P., 1960. Some characteristics of glauconite from the coastal plain formations of New Jersey. U.S. Geol. Surv. Prof. Pap., 400-B: 430--432.
283 Rogers, J., 1974. Surficial sediments and Tertiary limestones from the Orange-Liideritz shelf. SANCOR Mar. Geol. Prog. Tech. Rep., Geol. Dep. Univ. Cape Town, 6: 24--38. Ross, C.S. and Hendricks, S.R., 1945. Minerals of the montmorillonite group. U.S. Geol. Surv. Prof. Pap., 205-B: 23--79. Seed, D.P., 1965. The formation of vermicular pellets in New Zealand glauconites. Am. Mineral., 50: 1097--1106. Warshaw, C.M., 1957. The Mineralogy of Glauconite. Thesis, Penn. State Univ., University Park, Pa., 155 pp. (unpublished). Weaver, C.E. and Pollard, L.D., 1973. The Chemistry of Clay Minerals. Elsevier, Amsterdam, 213 pp.