The crystallographic fabric and texture of siderite in concretions: implications for siderite nucleation and growth processes

The crystallographic fabric and texture of sideritein concretions: implications for siderite nucleationand growth processes

MARK W. HOUNSLOW1

School of Environmental Sciences, University of East Anglia, Norwich NR3 7TJ, UK

ABSTRACT

The crystallographic fabric of siderite in siderite concretions has been

determined for upper Carboniferous (Westphalian-A) non-marine concretions

and lower Jurassic (Pliensbachian) marine concretions. Compositional zoning

indicates that individual siderite crystals grew over a period of changing pore

water chemistry, consistent with the concretions being initially a diffuse patch

of cement, which grew progressively. The siderite crystallographic fabric was

analysed using the anisotropy of magnetic susceptibility, which is carried by

paramagnetic siderite. The siderite concretions from marine and non-marine

formations exhibit differences in fabric style, although both display increases

in the degree of preferred siderite c-axis orientation towards the concretion

margins. The Westphalian non-marine siderites show a preferred orientation of

siderite c-axes in the bedding plane, whereas the Pliensbachian marine

siderites have a preferred orientation of c-axes perpendicular to the bedding.

In addition, a single marine concretion shows evidence of earlier formed,

inclined girdle-type fabrics, which are intergrown with later formed vertical

c-axis siderite fabrics. The marine and non-marine fabrics are both apparently

controlled by substrate processes at the site of nucleation, which was probably

clay mineral surfaces. Siderite nucleation processes on the substrate were most

probably controlled by the (bio?) chemistry of the pore waters, which altered

the morphology and crystallographic orientation of the forming carbonate. The

preferred crystallographic orientation of siderite results from the orientation of

the nucleation substrate. Fabric changes across the concretions partially mimic

the progressive compaction-induced alignment of the clay substrates, while the

concretion grew during burial.

Keywords Concretions, fabric, magnetic siderite, substrate, susceptibility.

INTRODUCTION

The fabric of sediments can be described througheither the grain shape fabric or in terms of thepreferred crystallographic orientation of mineralcrystallites, although these two fabrics are often

intimately interrelated. In particular there hasbeen much work on the origin of grain shapefabrics in carbonates, because fabrics are seen asimportant indicators of the origin of cements(Longman, 1980; Dickson, 1993). The crystallo-graphic fabric of cements in sediments is also afunction of their origin, because such fabrics arerelated to the chemical and physical environ-ment, which in¯uences crystal form (Given& Wilkinson, 1985; Jones & Kahle, 1993; Pedleyet al., 1996). Crystallographic and shape fabrics

1Present address: Geography Department, EnvironmentLancaster, Lancaster University, Bailrigg, LancashireLA1 4YW, UK (E-mail: [email protected]).

Sedimentology (2001) 48, 533±557

Ó 2001 International Association of Sedimentologists 533

also have the potential to indicate stress andstrain conditions during their formation, a featurewhich is widely utilized in metamorphic rocks(Etheridge et al., 1974; Tullis & Yund, 1982;Tarling & Hrouda, 1993). Fabric studies of diagen-etic minerals could equally yield informationabout the environmental constraints on the for-mation of these minerals, particularly if linked togeochemical and diagenetic histories. However,comparatively little is known about the crystallo-graphic fabric of diagenetic minerals.

Concretions are the segregation of cementsformed during diageneisis, mainly through theaction of organism stimulated reactions withinthe sediment (Raiswell, 1988; Curtis et al., 1986;Coleman & Raiswell, 1995). Concretion fabrics arerelatively simple systems to study because the dia-genetic histories of the component mineralscan be quite well constrained with geochemicalmodels (Mozley & Burns, 1993; Coleman & Rai-swell, 1995). This study examines the crystallo-graphic and grain shape fabrics of siderite insome siderite concretions and explores the factorsthat control the fabric characteristics. Concretionsfrom two settings are described: (a) concretionsand siderite concretionary sheets from non-marine sediments of Silesian age, and (b) concre-tions from marine mudrocks of Lower Jurassicage. The composition and mineralogy of theseconcretions are described together with the sid-erite crystallographic fabric, which has beenquanti®ed using the anisotropy of magnetic sus-ceptibility (AMS). Qualitative assessment of thegrain and crystallographic fabrics was also per-formed optically.

SAMPLES

The non-marine Westphalian A concretions sam-pled, form part of the deposits of a prodeltasequence at the level of the Alton (G. listeri) andParkhouse Marine Bands, which are exposed inquarries SE of Hepworth, South Yorkshire, UK(Table 1 and Fig. 1; Guion & Fielding, 1988). Thissequence has been extensively described in termsof its sedimentology (Guion & Fielding, 1988) andits clay and carbonate mineralogy (Ashby & Pear-son, 1979; Pearson, 1979, 1985). Various authorshave examined the carbonate concretions in thesesediments in terms of their geochemistry (Curtiset al., 1975, 1986; Love et al., 1983; Fisher et al.,1998) and compactional history (Oertel & Curtis,1972; Curtis et al., 1980). The samples examinedare from four concretionary horizons (Table 1).

Concretions IS25 and IS19 are from BullhouseQuarry (UK grid reference: SE209023) locatedabove the Alton marine band and are described interms of their geochemistry by Curtis et al. (1986).Siderite concretions HA and HB1 were collectedfrom Middlecliff Quarry, Hazelhead (Ashby &Pearson, 1979) and are located below the palae-osol of the Halifax Hard Bed Coal. All theconcretions are clay-hosted siderite concretions,numerous in the sequence. The exception is IS25,which is a siderite cemented sandy siltstone. Thenon-carbonate mineralogy of the concretions issimilar to the surrounding mudrocks and consistsof various clays (mixed-layered illite-smectite,kaolinite and smaller amounts of detrital chlo-rite), quartz and minor (<0á1%) pyrite (Ashby &Pearson, 1979; Curtis et al., 1986). Occasionallaminations indicate differential compactionaround the concretions, suggesting concretioncement formed before major compaction (Curtiset al., 1980, 1986). These sediments are unmet-amophosed, but have suffered minor folding andfaulting during basin inversion in the latestCarboniferous and early Permian (Lee, 1988).

The concretions from the marine sequence arefrom Lower Jurassic mudrocks in Crosby WarrenQuarry (Scunthorpe, north Lincolnshire, UK, gridreference: SE915128; Fig. 1) exposed by theformer quarrying of the Frodingham Ironstone(Gaunt et al., 1992; Taylor & Curtis, 1995). Sam-ples F5 and F6 are from the Lower Pliensbachian(Uptonia jamesoni ammonite zone), close to thePecten Ironstone (Table 1). Sample F6 is one-halfof a septarian concretion, in which the septa are®lled with a soft white clay. The matrix of the F6concretion subsampled did not have any penetr-ating septa. The sequence has suffered minorextensional faulting and post-Cretaceous tilting tothe east (Fig. 1; Gaunt et al., 1992).

METHODS

Elemental and mineralogical analysis

The mineralogy of the concretions was examinedusing X-ray power diffraction (XRD) combinedwith scanning electron microscopy (SEM) andback-scattered electron microscopy (BSEM).These observations and energy dispersive X-rayanalysis (EDXA) were performed on carbon-coated polished samples. Element concentrationsin siderite grains were determined from selectedpolished subsamples used for the fabric deter-mination, using a LINK Ltd, EDXA on a Hitachi

534 M. W. Hounslow

Ó 2001 International Association of Sedimentologists, Sedimentology, 48, 533±557

Table

1.

Su

mm

ary

of

sam

ple

san

dgen

era

lcon

cre

tion

data

.

Age/l

ocati

on

/sa

mp

lecod

eS

trati

gra

ph

icp

osi

tion

Con

creti

on

morp

holo

gy

Sid

eri

tem

orp

holo

gy,

gra

insi

ze

(vol.

%si

deri

te)

Sid

eri

tech

em

istr

y.

wt%

FeC

O3,

Mn

CO

3,

CaC

O3,

MgC

O3

Refe

ren

ce

West

ph

ali

an

-non

-mari

ne,

Hep

wort

h,

S.

York

shir

eIS

25

5á8

mabove

Park

hou

sem

ari

ne

ban

dC

on

creti

on

ary

sheet

in,

san

dy

silt

ston

eS

ubh

ed

ral,

~5±10

lm(5

±58%

)83á6

,1á5

,6á7

,7á0

*82á3

,1á3

,2á6

,13á6

C

urt

iset

al.

(1986)

IS19

1á9

mabove

Hali

fax

Hard

Bed

Coal

Con

creti

on

ary

sheet

Len

gth

-fast

,elo

ngate

,~2

5lm

size

(52%

-87%

)81á3

,11á4

,5á2

,2á3

82á1

,4á9

,2á9

,9á6

C

urt

iset

al.

(1986)

HA

1á5

mbelo

wgan

iste

rof

Hali

fax

Hard

Bed

Coal

Nod

ula

rcon

cre

tion

ary

sheet

Len

gth

-fast

,elo

ngate

,~2

5lm

size

(51%

-88%

)80á5

,14á0

,4á8

,0á7

Th

isst

ud

y

HB

12

mbelo

wgan

iste

rof

Hali

fax

Hard

Bed

Coal

Obla

ten

od

ule

Len

gth

-fast

,elo

ngate

,~2

5lm

size

(ND

)N

DT

his

stu

dy

Jura

ssic

-mari

ne,

Cro

sby

Warr

en

,L

incoln

shir

eF

610

mabove

Pecte

nIr

on

ston

eS

ligh

tly

sep

tari

an

obla

tecon

cre

tion

Su

bh

ed

ral,

~10

lm(8

1±86%

)68á3

,1á3

,11á8

,18á6

Th

isst

ud

y

F5

2m

belo

wP

ecte

nIr

on

ston

eC

on

creti

on

ary

sheet

Rh

om

boh

ed

ral,

~10

lm(4

7±64%

)65á9

,2á9

,15á9

,15á3

Th

isst

ud

y

Th

elo

cati

on

sof

the

Hep

wort

hqu

arr

ies

are

deta

iled

inA

shby

an

dP

ears

on

(1979)

an

dF

ish

er

etal.

(1998),

an

dC

rosb

yW

arr

en

inG

au

nt

etal.

(1992).

*A

naly

sis

for

layer

1on

ly,

no

data

=N

D.

Data

of

Cu

rtis

etal.

(1986).

Siderite fabrics 535


S450 SEM, with an activation volume diameter ofabout 5 lm, with ZAF and matrix corrections(Reed, 1996). Grains for this analysis were selec-ted only if they were larger than 5 lm, with thebeam positioned at the grain centre. EDXAanalyses of metal oxide <0á1% are taken as zero(detection limit 0á1±0á5% of metal oxide). Fe, Mg,Ca and Mn are expressed as molar carbonatefractions. Siderite abundance was determinedwith BSEM digital images of the polished surfa-ces (at magni®cation ´300), as described by Pye(1984). The images were separated into siderite,pyrite and calcite + clay + quartz + porosity win-dows using the LINK software. This operation isrelatively simple because of the distinct differ-ence in the back-scatter intensity of these threemineral groupings. As a result of back-scatter gaindrift and edge effects, siderite abundances areestimated to be accurate to within �1%. Theseareal abundances are assumed to be approxi-mately equivalent to siderite volumetric abun-dance. The fabric and textural features of theconcretions were also qualitatively investigatedusing ultra-thin sections.

Fabric: anisotropy of magneticsusceptibility (AMS)

AMS is the directional variation in magneticsusceptibility and has been extensively used inassessing preferred grain orientation in sedimen-tary, metamorphic and igneous rocks (Rochette

et al., 1992; Tarling & Hrouda, 1993). The AMS isconveniently represented by a susceptibilityellipsoid which has three orthogonal principalaxes, the maximum, minimum and intermediatesusceptibility axes (Kmax, Kmin, Kint respectively).The shape of this ellipsoid is represented by theq-value (possible range 0 for fully oblate to 2á0 forfully prolate), and the degree of anisotropy of theellipse by h% where: q � (Kmax ± Kint)/[(Kmax +Kint)/2] ± Kmin; h% � 100 ´ (Kmax ± Kmin)/(Kmax +Kmin)/2.

Cubic samples (edge length 1 cm or 1á5 cm) forAMS determination were cut from a sectionedslice through the thickest part of the concretionsHA, F5 and F6. The AMS of these was measuredon a MOLSPIN anisotropy instrument (Hounslowet al., 1988). In order to evaluate the verticalheterogeneity of the fabric in concretion HA, anadditional set of specimens (8 mm thick by25 mm diameter, denoted sample set HA-1) werecut 20 mm from the cubic sample section locatedthrough the concretion centre. This slice-methodof sampling was used on concretion HB1. Theseslice specimens were measured on a high ®eldtorque magnetometer (Hounslow, 1984; Tarling& Hrouda, 1993), using a magnetic ®eld of 0á36T.The AMS of the cubic-samples from IS19 werealso measured using a high ®eld torque magne-tometer indicating that the q and directionalresults are equivalent to the MOLSPIN measure-ments (Hounslow et al., 1988). The h% valuesderived from the high ®eld torque magnetometer,

Fig. 1. Location of the study sites and simpli®ed geology of part of eastern England.

536 M. W. Hounslow


and the MOLSPIN units, are not directly compar-able because of the different magnetic ®elds used(Hounslow, 1984). Repeat measurements andmeasurements on different instruments indicateq and h% errors are typically �0á01 and �0á1,respectively, for the specimens described here(Hounslow, 1984).

The interpretation of the AMS depends onwhich minerals carry the fabric (Hrouda, 1980;Rochette et al., 1992), hence the magnetic min-eralogy was investigated. The ferrimagneticmineralogy of the samples was investigatedusing isothermal remanent magnetic measure-ments up to 1T (Rochette, 1987; Tarling &Hrouda, 1993). The minerals responsible forthe susceptibility have been investigated usinga Bartington Ltd low temperature magneticsusceptibility meter using temperatures between77 °K and 290 °K. (Richter & Van der Pluijm,1994). The minerals responsible for the AMSwere investigated using the torque-magnetic®eld behaviour on a high ®eld torque magne-tometer, using ®elds up to 0á8T (Owens &Bamford, 1976; Hounslow, 1986).

RESULTS

Hepworth: Westphalian non-marineconcretions

Chemistry and mineralogical analysis

The concretions IS19, HA and HB1 are largelycomposed of siderite, with a siderite mean grainsize of about 25 lm. Rare grains of subhedralpyrite and small amounts (<5%) of clay min-erals can be found enclosed within siderite(Fig. 2). In the concretion centres, siderite hasan abundance of up to 88%, reducing to about55% at the concretion margins (Figs 3 and 4;Table 1; Curtis et al., 1986). Mn, Ca and Mg allshow signi®cant substitutions for Fe in thesiderites (Table 1; Figs 3 and 4). Compositionalzoning of the siderite grains is shown by theBSEM images and X-ray mapping (Fig. 2A andC). The X-ray maps indicate some grain centreshave an elongate shape and are signi®cantlyenriched in Mn (Fig. 2D). Ca and Mg in thesiderite are preferentially concentrated in thegrain periphery, although overall Mg is appar-ently most concentrated in the interstitial clays(Fig. 2A and C). These grain textural and com-positional features are similar to those noted byFisher et al. (1998) from other similar sideriteconcretions in the Hepworth sediments.

The EDXA spot analyses show much largervariation between siderites at the same position,than between different positions across the con-cretions (Figs 3 and 4). However, systematicchanges are indicated in IS19, which has higherMn/Fe and lower Mg/Ca in the concretion centrethan the periphery. Mg/Fe and Ca/Fe also haveelevated values at the upper margin in compar-ison to layer 3 of concretion IS19. The Mg/Fe ratiois also lower in layer 4 (centre) of HA than themore peripheral parts, whereas any other system-atic trends in this concretion are masked byadjacent grain variability. EDXA analyses alsoindicate small amounts (<0á4 wt% metal oxide) ofSi, Al and P (frankolite?; Fisher et al., 1998), fromclay and quartz inclusions.

The textural features of IS25 are quite differentfrom the other Hepworth concretions. IS25 is asiderite cemented sandy siltstone with a widevertical variation in siderite abundance (Table 1;Curtis et al., 1986). The siderite mostly occurs inpatches up to about 100 lm in size, which arecomposed of smaller (�5±10 lm) subgrains(Fig. 2F). These patches are about the same sizeas the sand/silt grains, indicating siderite may bea replacement of former detrital grains. Small,�5 lm siderite crystals are also scattered throughthe clay and quartz matrix between the largersiderite patches. Clusters of siderite crystals inIS25 are also relatively enriched in Mn incomparison to adjacent grain clusters. Theseclusters display changes in back-scatter intensityprobably attesting to some zoning and intergraincompositional differences (Fig. 2E and F). Grainsin IS25 are signi®cantly enriched in Ca and Mgand depleted in Mn in comparison to IS19(Table 1), which is a similar trend to the bulkchemical data of Curtis et al. (1986).

The Mn-enriched, and Ca + Mg enriched, por-tions of crystals most probably match two sideritetypes previously found in these concretions basedon 1014 peak separation on XRD traces (Pearson,1974, 1985; Curtis et al., 1975). Curtis et al.(1975) suggested one of these phases was rich inMn and Fe whereas the second one was richer inMg and Ca, and largely restricted to the outer edgeof the concretion. Based on the data here and inFisher et al. (1998) these two end-member side-rites are in fact present throughout the width ofthe concretion bodies.

Fabric

Ultra-thin sections of IS19, HA and HB1 perpen-dicular to the bedding show an apparent mixture



Fig. 2. Siderite concretions from the Hepworth section. X-ray maps (A, C and E) and BSEM images (B, D and F) at thesame position. The X-ray maps were computer enhanced using local median and averaging ®lters. The BSE imageswere cropped slightly smaller (mostly off the base) than the corresponding X-ray maps. (A) IS19: X-ray maps of thesite shown in B. Note the enrichment of Mg (arrows) around the darker rims in the BSEM image (pointers). Mn is alsoenriched in some crystal centres (arrow in A). (C) HA: layer 5, X-ray maps of the site shown in D. The bright phase inD is pyrite (pointer), and the earlier formed centres of grains are enriched in Mn, with a peripheral concentration ofCa (arrows in C). Mg is most concentrated in the clay fraction, but may show some concentration in the outer-mostedges of the crystals (arrows). (E) Sample IS25: X-ray maps of the position shown in F. The siderite grains showdifferences in back-scatter intensity across subgrains (pointer) perhaps re¯ecting patchy compositional differences inMn and Ca (arrow in E).

538 M. W. Hounslow


of rounded and elongate grains due to a preferredshape orientation, with the long axis parallel tothe bedding plane (Fig. 5A). Use of a quartzwedge shows the elongate siderite grains arelength-fast, hence with the c-axis parallel to thelength (Fig. 5A). The majority of these grainshave even extinction, except for occasionalelongate grains with length-parallel twinningand some with 2±3 subregions, with differentextinction angles. This suggests that for the mostpart the siderite grains grew progressivelythrough syntaxial overgrowths on the priorformed crystals. Siderite crystals in IS25 aremore subhedral in shape and as far as can beevaluated for such small grains, appear to displayeven extinction.

The mass susceptibility of the samples fromconcretions IS19, IS25, HEP1 shows strongacross-concretion differences, which are relatedto the abundance of siderite (Figs 3, 4 and 6). Theh% values vary vertically across all the concre-tions, with low h% values in the centre andhigher values at the margin (Figs 3, 4 and 6;Table 2). IS25 is an exception to this and displayslower h% values near the margins (Table 2). Theh% data for HA-1 and HB1 is less regular acrossthe concretions (Fig. 6). This is a re¯ection ofheterogeneity in the AMS fabric, and the fact that

the data for IS19, IS25 and HA has been smoothedby averaging several specimens (Table 2).

The susceptibility axes show an oblate mag-netic fabric (q < 0á1) with the Kmax axes distri-buted about the horizontal plane, and the Kmin

near vertical (Fig. 7 and Table 2). The AMS ofthese samples is nearly symmetrically uniaxial inform, but the clustering of Kmax indicates thefabric shows signi®cant but weak preferred Kmax

orientation within the bedding plane (Table 2 andFig. 7). The anisotropy within the horizontalplane is responsible for the differences in thebehaviour of the q-value between IS19, HA andHB1. A stronger horizontal lineation is present inthe centre of IS19 compared with the periphery,whereas the reverse situation occurs for HA, IS25and HB1 (Table 2).

Crosby Warren, Jurassic marine concretions

Chemistry and mineralogical analysis

Both F5 and F6 have similar mineralogy consistingof siderite, calcite, quartz and clays (illite andkaolinite) and small amounts of pyrite (Fig. 8). Thenon-carbonate mineralogy is similar to the sur-rounding mudrocks (Taylor & Curtis, 1995). Thecalcite is present in the concretions as biogenic

Fig. 3. Chemistry and AMS data for Hepworth, non-marine concretion IS19. Each of the molar ratio symbols(MgCO3/CaCO3, MgCO3/FeCO3 MnCO3/FeCO3 and CaCO3/FeCO3) represents a spot analysis (two to four per posi-tion) at a corresponding level on a specimen from the layer indicated. The full and dashed lines represent theaveraged compositional ratios at each vertical position. The q, h% and susceptibility values are the average of severalspecimens from each layer (Table 2).



fragments, which may sometimes show evidenceof replacement with siderite. The average sideritegrain size is about 10 lm. The siderite abundancesfor F6 and F5 are 85% and 55%, respectively, andin the case of F5 show a 15% decrease at the bottomedge of the concretion, in comparison to the centre(Fig. 9). Siderite abundance in F6 shows a �2%increase at the horizontal margin of F6, and small¯uctuations (�5%) in the vertical variation inabundance (Fig. 10). Such variations may be rela-ted in some way to textural variations in theoriginal host sediment, because burrows in thehost sediment have clearly perturbed the sideriteabundance (Fig. 10).

The marine siderites are signi®cantly richer inMg and Ca than the non-marine Westphalianconcretions (Table 1 and Fig. 11). In concretionF6, BSEM shows two siderite phases with acompositional boundary; a later formed phase,which is brighter, and coats an earlier formedphase, with a more prominent rhombic grainshape (Fig. 8C). The intergrowth of these twophases is on too small a scale (i.e. <5 lm) forX-ray mapping to clearly show compositionalchanges across grains. However, clusters of theearlier formed grains, in X-ray mapping, indicatethat it is depleted in Ca compared with the laterformed phase (Fig. 8A and B). The siderite grains

Fig. 4. Chemistry and AMS data for Hepworth, non-marine concretion HA. Chemical analysis and percentage sid-erite were not determined for layers 1, 2 and 3. Symbols as in Fig. 3.

540 M. W. Hounslow


in concretion F5 appear to be homogeneous whenexamined under BSEM and X-ray mapping,although a similar spread in chemical composi-tion is apparent for both F5 and F6 (Figs 9 and10). EDXA analyses also indicate small amounts(<0á4 wt%) of Si, Al and P. The Si and Al isprobably from clay and quartz inclusions.

Like the Hepworth concretions, the adjacentgrain chemical variability is mostly larger than anycentre-periphery changes (Figs 9 and 10). How-ever, there is indication that the lower margin of F5has larger Mg/Ca and lower Ca/Fe than the concre-tion centre. Concretion F6 also shows systematicchanges in Ca/Fe and Mg/Ca (vertical traverse), notrelated to centre-periphery changes, but to ¯uctu-ations in the amount of siderite in the lower part ofthe concretion (Fig. 10A). Detailed XRD examina-

tion of the siderite 1014 peak in concretions F5 andF6 indicates a subsidiary peak at 2á785AÊ (�0á003)on the dominant d-spacing at 2á807AÊ (�0á001). Thisprobably indicates that siderites in both F5 and F6are also dual-phase, but the texture of this inter-growth is not apparent on BSEM images of F5.These XRD peak positions are similar to thosefound by Pearson (1985) from the Hepworthsiderites, with peaks centred at 2á800AÊ (�0á009)and 2á785AÊ (�0á001).

Fabric

The shape texture of siderite grains in sample F6 ismore subhedral than that of F5 which are typicallyrhombohedral (Fig. 8). As far as can be evaluatedfor such small grains, both F5 and F6 appear toposses even extinction throughout each grain.

Fig. 5. C-axis and grain long axisorientation for Westphalian non-marine concretions based on thin-section data. (A) HB1 (19 mm fromconcretion margin). (B) HA-1 (41 mmfrom concretion margin). (C) Sum ofthe c-axis orientation data for HB1and HA-1. The mean (black arrows)and number of grains measured (N)are indicated. Sector angular interval10°.

Fig. 6. AMS data for Hepworth non-marine specimens HA-1 and HB1.These represent individual samplemeasurements, rather than averageddata as displayed in Figs 3 and 4.Arrows indicate the location of thethin-section data shown in Fig. 5.



The mass susceptibility of the samples fromconcretion F5 show differences across the con-cretion which are related to the abundance ofsiderite (Fig. 9 and Table 3), whereas concretionF6 does not show such marked mass susceptibil-ity differences because the siderite abundance isrelatively constant. As in the case of the Hep-worth concretions, the h% values vary verticallyacross both F5 and F6, with low values in thecentre and higher values at the margin (Figs. 9and 12; Table 3). However, F6 shows a morecomplex relationship when examined in a hori-zontal direction through the concretion centre(Fig. 12).

The Crosby Warren concretions possess avertical Kmax fabric with a dominantly prolateellipsoid (Figs 9 and 12; Table 3), quite unlike theHepworth concretions (Fig. 7). The samples fromF5 have q-values approaching 2 indicating anearly symmetrical uniaxial fabric (Table 3),which is also re¯ected in the tight cluster of Kmax

axes (Fig. 7). The sample coverage in concretionF6 indicates a more complex fabric than simplecentre-periphery changes. It shows a horizontalchange in the h% and q-values and also differ-ences in the vertical variation at different hori-zontal positions away from the concretion centre.The q-values in the approximate centre of the

Table 2. Summary of mean AMS data for the Hepworth concretions.

Kmax Kminv

Layer Dec, Inc s3 s2 R Dec, Inc s3 s2 R N h% q (´10±7 m3/Kg)

Sample HA1 124, )5 3á96 0á04 3á98 305, )84 3á98 0á01 3á99 4 13á5 0á05 4á772 125, )3 3á70 0á30 3á85 353, )85 3á99 0á00 3á96 4 10á9 0á04 5á973 123, )2 2á58 0á42 1á21 007, )87 3á00 0á00 3á00 3 9á08 0á03 7á284 186, )4 3á08 0á92 2á87 044, )85 4á00 0á00 4á00 4 7á05 0á02 7á815 250, )10 4á91 0á08 4á96 070, )79 5á00 0á00 5á00 5 5á35 0á04 8á456 257, )12 4á93 0á06 4á97 081, )77 4á98 0á02 4á99 5 8á72 0á05 7á697 135, )7 3á93 0á07 3á97 315, )83 3á99 0á00 4á00 4 12á1 0á05 6á84

Total 289, )2 20á1 8á72 13á9 041, )85 28á5 0á43 28á7 29 9á53 0á04 6á97

Sample IS251 248, )5 5á65 0á27 4á00 031, )81 5á72 0á19 5á86 6 0á62 0á23 6á242 259, 0 4á17 0á82 1á07 355, )87 4á97 0á02 4á99 5 4á00 0á09 2á423 228, )1 3á99 0á01 0á11 342, )89 3á96 0á04 3á98 4 1á59 0á13 4á20

Total 245, )2 13á3 1á62 4á37 022, )86 14á6 0á25 14á8 15 2á07 0á15 4á29

Sample IS191 317, )8 2á36 0á59 1á56 112, )80 2á83 0á15 2á91 3 11á9 0á04 6á912 340, 0 1á94 0á95 1á14 303, )83 2á78 0á21 2á89 3 4á75 0á08 9á813 122, 0 2á48 0á51 0á82 351, )85 2á93 0á05 2á96 3 6á41 0á03 7á88

Total 315, )3 6á46 2á30 3á46 316, )84 8á54 0á38 8á76 9 7á67 0á05 8á20

*Sample HA-11±4 100, )6 3á92 0á07 3á96 67, )79 3á98 0á01 3á99 4 22á1 0á05 3á945±8 101, )6 3á69 0á19 2á13 62, )78 3á97 0á03 3á98 4 19á7 0á05 4á229±12 93, )14 2á86 0á96 3á23 43, )86 3á98 0á02 3á99 4 18á1 0á02 4á35

Total 99, )8 10á4 1á17 9á23 62, )81 11á87 0á10 11á93 12 19á9 0á04 4á17

*Sample HB11±4 33, )25 2á08 1á86 2á89 313, )87 3á99 0á01 3á99 4 24á5 0á04 3á235±8 160, )2 3á86 0á14 3á93 88, )89 4á00 0á00 4á00 4 16á3 0á03 5á149±12 151, )2 3á08 0á91 3á43 51, )89 3á99 0á01 3á99 4 17á9 0á02 5á03

Total 161, )4 8á71 2á80 6á82 342, )89 11á96 0á03 11á98 12 19á6 0á03 4á47

The layer numbers correspond to the sample layers/samples depicted in Figs 3, 4 and 6, numbered consecutivelyfrom the top. Axial directions are relative to an arbitrary horizontal direction for IS19 and IS25, and relative to northfor HA, HA-1 and HB1. The mean of the AMS axes were determined using eigen statistics (Mardia, 1972). s3 and s2

are the two largest eigenvectors, and R the resultant length. N is the number of specimens in each layer used tocalculate the means. *High ®eld torque measurements.

542 M. W. Hounslow


concretion indicate that the fabric has a moretriaxial form, with q � 0á4 in the centre-mostsample. This AMS changes into a strongly prolateellipsoid at the margin, particularly the horizon-tal margin (Fig. 12). The vertical and horizontalmargins show different q and h% values, indica-ting the fabric is distinctly different.

Mineralogical source of the AMS

In these sideritic concretions ferrimagnetic min-erals are not responsible for changes in suscepti-bility and AMS, which appear to be mostlycarried by siderite. The evidence supporting thisis:

1 The susceptibility variation (v) with absolutetemperature closely follows the Curie±Weiss law,v � C/(T ) h), where C is the Curie Constant, T isthe absolute temperature and h is the paramag-netic Curie Temperature (Fig. 13A). There is noindication of superparamagnetic, single domainor multidomain ferrimagnetic particles, whichcan be detected by deviation of the v vs. 1/Trelationship from paramagnetic linearity (Franceet al., 1999). This indicates the susceptibility iscarried entirely by paramagnetic minerals (Roch-ette, 1987; Richter & Van der Pluijm, 1994).

2 The isothermal remanent magnetization indi-cates saturation by �0á3T, indicating a ferrimag-net (i.e. magnetite) or a magnetic sulphide carriesthe remanent magnetic properties. However, thisferrimagnetic mineral is evidently in such smallquantities that it makes no signi®cant contribu-tion to the susceptibility.

3 The high ®eld magnetic torque due to theanisotropy does not show saturation at ®elds upto 0á7T (Fig. 13A), which indicates the AMS iscarried by a paramagnetic mineral (Owens &Bamford, 1976). Ferrimagnetic or antiferrimag-netic contributions to the AMS lead to saturationin the magnetic torque-applied ®eld relationship.

4 Siderite is the major constituent of theconcretions, constituting a weight per cent of88% to 62%, for a volume per cent of 85% and55%, and is also the major repository of Fe.Siderite also has one of the largest susceptibilities(mass susceptibility, 11 ´ 10±7 m3/Kg; Kristmanet al., 1932; Powell & Miller, 1962; Jacobs, 1963)and crystalline anisotropy of any commonlyoccurring paramagnetic mineral (Rochette et al.,1992). Pyrite has a small isotropic magneticsusceptibility (Rochette, 1987) and, therefore,will not contribute to the AMS. Chlorite is knownto be an important carrier of the AMS in some

Fig. 7. Equal area stereographicprojection of Kmax (squares) andKmin axes (triangles) for concretionsF6, F5, HA, IS19 and IS25. The datafor HA are oriented relative to north(N), whereas all others are relative toan arbitrary horizontal referencedirection.



mudrocks (Hounslow, 1986; Rochette, 1987) andoccurs as detrital grains in both the Hepworth(2±4 wt% of the silicate mineralogy; Ashby &Pearson, 1979) and the Crosby Warren clays(Taylor & Curtis, 1995). Therefore, it is possibleFe-bearing clays may make a minor contributionto the AMS.

INTERPRETATION OF THE CONCRETIONMINERALOGY/CHEMISTRY

The bulk chemical data of Curtis et al. (1986) andFisher et al. (1998) demonstrate a consistentpattern of centre-periphery siderite substitutionchange in the Hepworth concretions, with thecentral parts of the concretions generally having

lower Mg/Fe, larger Mn/Fe ratios and moreenrichment in 13C. Curtis et al. (1986) explainedthe bulk siderite geochemistry of the Hepworthconcretions in terms of a concentric concretiongrowth model concomitant with pore water evo-lution and progressive burial. These authorsexplain the phenomenon of Mn and 13C enrich-ment in concretion centres as a result of Mn(IV)oxidation of organic matter in combination withmethanogenic processes favouring carbonate pre-cipitation enriched in Mn and Fe in earlydiagenesis. This process was replaced progres-sively in time by decarboxylation and enrichmentwith Mg (Coleman, 1985). Fisher et al. (1998)have refuted this model for the Hepworth con-cretions, suggesting the concretions may havegrown entirely within the methanogenesis zone atdepths less than 10 m.

Fig. 8. X-ray maps (A) and BSEM images (B, C and D) of Jurassic marine siderite concretions. X-ray maps asdescribed in caption to Fig. 2. (A) sample F6, layer 3: X-ray maps of the site shown in B (BSEM cropped slightlysmaller). The ®rst formed and later brighter phase are intimately intergrown. The cluster of darker, earlier formedsiderite (pointer) is depleted in Ca (arrow in A) compared to the surrounding mixture of phases. (C) close-up of theintergrowth of the two phases in sample F6. (D) Brighter siderite rhombs (pointer) in F5 interlocking with darkerquartz, calcite and clay minerals in a slightly porous texture.

544 M. W. Hounslow


The analyses here and in Fisher et al. (1998)indicate that crystals are both strongly zoned, andcan display signi®cant chemical differencesbetween adjacent grain centres. The centre±edgecrystal composition changes measured by Fisheret al. (1998) are for Mg, 1:2á5±1:>30; for Mn, 1:1 to�20 : 1 and for Ca, 1:1á7±1:4. These indicate thatadjacent crystals can display large differences intheir centre-edge composition. If it were assumedcomposition was related to time of formation,these data could be inferred to suggest adjacentcrystals potentially nucleated at different timesand grew at different rates.

In IS19, the bulk data of Curtis et al. (1986)show centre to periphery changes in Mg/Fe from0á05 to 0á16, Mn/Fe from 0á045 to 0á07 and Mg/Cafrom 2 to 4á1. The crystal-centre bias provided bythe spot analyses indicates that on average in theconcretion centre, the crystal centres are deple-ted in Mg 2á4-fold and enriched in Mn and Ca bya factor of 2á7 and 1, respectively, in comparisonto the average grain composition. For the con-cretion edges, the crystal centre±bulk depletionfactor is 6á2 for Mg and enrichment in Mn andCa by 1á9 and 1á6 respectively. What is not clear

is how much of the across-concretion bulkchanges are attributable to either changes in thechemistry of the crystal zoning (growth rate?),or to the difference in abundance of theMn-enriched crystal centres. Fisher et al. (1998)suggest �5% more of the later Mg-enrichedsiderite could account for the bulk across-con-cretion changes.

These data indicate initially dispersed sideritecrystallization throughout the concretion body,which was followed progressively by more dis-persed crystallization together with overgrowthson earlier grains, with an evolution to Mgenriched and 13C depleted siderite. Therefore,the most likely model for these concretions is acombination of the diffuse model of Mozley(1996), which accounts for most of the growth,but with some preference for later cement forma-tion (new crystals centres and/or more over-growth) in the outer parts of the concretion. Aprogressive burial model during siderite growth ismore compatible with the changes in sideritefabric across the concretions discussed below.

The compositional complexity of concretionsiderite has also been described by Mozley

Fig. 9. Chemistry and AMS data for Crosby Warren concretion F5. Chemical spot analyses and percentage sideritewere not determined for layer 1. See Fig. 3 for details. The drop in % siderite at 18 mm is due to the presence of 14%pyrite.



(1989a) who showed changes in the relativeabundance of two siderite phases, from concre-tion centre to periphery. Compositional zoning ofsiderite crystals and similar across-concretionvariation have also been found by Matsumoto& Iijima (1981) in Tertiary coal-bearing sedi-ments, suggesting this texture may be widespreadin non-marine siderites.

The Crosby Warren marine siderites are deple-ted in Mn, enriched in Ca and Mg and have lower

Mg/Ca ratios compared with the non-marinesiderites (Fig. 11 and Table 1). These composi-tional differences are compatible with the moreextensive compositional data of Mozley (1989b),who suggested such differences are due to thecontrasting ionic contents of fresh and sea water.The abundance of SO4 in marine derived pore-water means that pyrite is the ®rst major sink forFe (Gautier, 1982; Allison & Pye, 1994), account-ing for its early (i.e. framboidal) presence in the

Fig. 10. Chemistry of Crosby War-ren F6 concretion in both verticaland horizontal traverses. Traverselocations are indicated in Fig. 12.The AMS sample locations relativeto the spot analyses are also indi-cated. See Fig. 3 for details. Thedrop in %siderite in the horizontaltraverse at 43 mm is associatedwith a burrow ®ll. Chemical ana-lyses were only performed onhorizontal traverse samples shown.

546 M. W. Hounslow


sediment. For siderite to form in sediments frommarine pore waters, SO4 must have been exhaus-ted, leading to carbonate generation by methano-genesis (Gautier, 1982; Coleman, 1985; Spears,1989). If abundance of siderite is taken to beindicative of sediment porosity then the initialsiderite formed earlier in concretion F6 than inF5 ± this is supported by the fabric data discussedbelow. The dual phase siderite in the marineconcretions is more compatible with a diffuseconcretion model than a concentric growthmodel, although the fabric discussed below iscompatible with progressive burial and sideritegrowth.

INTERPRETATION OF THECONCRETION FABRIC

The investigation of the mineralogy indicates thatthe paramagnetic mineral siderite is responsiblefor both the susceptibility and AMS of theseconcretions. If the susceptibility is carried byparamagnetic minerals, any resultant AMS is dueto the preferred crystallographic orientation of theminerals, whereas a collection of grains withcompletely random crystallographic orientations,would theoretically possess no AMS (Richteret al., 1993). Few workers have examined thecrystalline susceptibility anisotropy (i.e. relatedto crystal structure) of paramagnetic minerals,consequently little is known about its variation innatural minerals other than the simplest detailsand inferences based on crystal structure (Krist-man et al., 1932; Rochette, 1987). Siderite isunusual because it possesses a very strong crys-talline anisotropy (Kmax/Kmin � 4á5; Jacobs, 1963;Rochette et al., 1992 quotes lower value) andlarge susceptibility (Kristman et al., 1932; Roch-ette, 1988). The anisotropy of siderite is essen-tially uniaxial with the maximum susceptibility(Kmax) directed parallel to the c-axis (similar toother common carbonates, e.g. Kristman et al.,1932; Rochette, 1988).

Consequently, as a result of crystalline anisot-ropy a group of siderite crystallites will have acollective maximum susceptibility (Kmax) direc-ted parallel to the preferred c-axis orientation,and the minimum susceptibility (Kmin) in a planenormal to this. Quantitatively, for a rock in whichthe density distribution function of crystallo-graphic c-axes can be described as f(u,w), thesusceptibility tensor of the rock, k, can beexpressed as follows (Hrouda, 1980):

k �Zu

Zw

f �u;w�K�u;w� sinw:du:dw

where K(u,w) is the anisotropic susceptibility ofthe crystal, and u,w are the polar angles specify-ing the orientation. With progressively largerdegrees of preferred c-axis orientation, the h%value will become larger. The q-value can beequated to the form of the distribution of c-axes,with q � 0á66 (oblate fabric) if the c-axes aredistributed in a plane and q � 0á66 (prolatefabric) if the axes are centred about an axialdirection.

For the Crosby Warren marine concretions, theAMS suggests a vertical preferred orientation ofsiderite c-axes such as that in Fig. 14A. This styleof AMS fabric is common in siderite and Fe-dolo-mite bearing marine sediments (Hounslow, 1984;Rochette, 1988; Rochette et al., 1992). This formof crystal fabric is similar to X-ray texturemeasured siderite fabric in the Mans®eld MarineBand (at the Westphalian B/C boundary; Attewellet al., 1969). A similar dolomite c-axes fabric hasalso been described from the Permian Marl Slate(Attewell et al., 1969) and from petrographicalobservation on marine dolomites, calcites andsiderites (Bellamy, 1977, 1980). In contrast, theHepworth non-marine concretions possess anoblate magnetic fabric, indicating a preferredorientation of siderite c-axes in the beddingplane, such as that depicted in Fig. 14B.

The AMS fabric in the Crosby Warren marineconcretion F6 is more complex and suggests thesuperimposition of two fabric types, (a) a subhori-zontal siderite c-axis preferred orientation, pre-valent in the concretion centre, and (b) a verticalsiderite c-axis preferred orientation, prevalent inthe periphery with larger q-values. However, thelower h% values in the centre of the horizontaltraverse (Fig. 12) are not entirely compatible withthe AMS being caused by a transitional fabric. Amix of these fabric types at exactly 50:50 wouldtheoretically produce zero h%. This suggests amore complex, possibly non-axial fabric, whichcannot be resolved by the AMS data, and isdiscussed later after examination of the processeswhich could give rise to siderite c-axis preferredorientation.

Mechanisms of fabric production

A tectonic, or direct compactional, origin forthese siderite fabrics is unlikely. First, because



the Westphalian and Jurassic are essentiallyundeformed. Second, a variety of carbonate fab-rics (prolate, oblate and other styles) are knownfrom both the Westphalian and the Jurassic(Attewell et al., 1969; Bellamy, 1980). Third,deformation of the surrounding shales in theHepworth concretions indicates the concretionsformed precompaction (Curtis et al., 1980, 1986),and the chemistry of the siderite indicates forma-tion during early diagenesis (Fisher et al., 1998).These data indicate that the siderite fabric musthave been generated during crystal growth. Inorder to understand these fabrics, two topics arediscussed, (a) the likely growth process of thesiderite crystallites, and (b) how this growthprocess could give rise to preferred crystallo-graphic orientation.

Siderite growth processes

Evidence from the Hepworth concretions indicatesthat the siderite formed syntaxial overgrowths oncompositionally different, earlier-formed grains.This distinction is not so clear for the Jurassicconcretions because of the ®ner grain size, but therim coating character of the brighter sideritephase in F6, and optical continuity of thesegrains, indicates that the siderite is also mostlya syntaxial overgrowth. Syntaxial overgrowths arelikely to be favoured, from a kinetic standpoint,because such growth sites are energeticallyfavourable (Berner, 1980; Sunagawa, 1994). Con-sequently, preferred crystallographic orientation(PCO) produced during crystal growth couldeither be controlled by the orientation of the

Fig. 11. FeCO3-CaCO3-MnCO3-MgCO3 ternary diagrams of siderite composition (weight percentage, apices markedFe, Ca, Mn or Mg). (A) Hepworth non-marine concretions (i) � sample HA (ii) � Samples IS19 and IS25 (®lledcircles � IS19 and open circles � IS25). (B) Jurassic marine concretions (i) � sample F6 and B (ii) � sample F5.

548 M. W. Hounslow


Table 3. Summary of mean AMS data for the marine siderite Concretions F5, F6.

Kmax Kminv

Layer Dec, Inc s3 s2 R Dec, Inc s3 s2 R N h% q (´10±7 m3/Kg)

Sample F61 176, )84 3á96 0á03 3á98 275, 0 3á95 0á03 0á26 4 10á9 1á26 6á812 105, )82 5á83 0á10 5á92 275, )6 5á48 0á47 1á86 6 5á67 1á28 6á993 116, )81 7á71 0á23 7á85 268, )2 5á51 2á28 3á14 8 4á08 1á01 6á924 125, )83 7á71 0á27 7á85 272, )3 5á45 2á54 5á24 8 4á52 1á04 6á985 070, )87 5á93 0á06 5á96 287, )3 4á41 1á58 3á69 6 6á71 1á19 6á876 092, )87 4á98 0á01 4á99 271, )2 4á95 0á05 2á99 5 9á07 1á07 6á88

Total 116, )84 36á0 0á71 36á5 274, )3 29á5 7á32 15á4 37 6á83 1á14 6á91

Sample F51 088, )86 3á98 0á02 3á99 227, )2 3á32 0á80 1á55 4 6á07 1á83 5á162 101, )85 3á98 0á01 3á99 218, )2 3á52 0á41 2á00 4 3á71 1á61 5á693 044, )86 4á00 0á00 4á00 198, )3 3á77 0á23 1á89 4 6á89 1á85 6á064 327, )87 3á99 0á00 4á00 189, )2 3á83 0á17 3á91 4 10á51 1á84 5á78

Total 069, )87 15á9 0á06 16á0 205, )2 13á6 2á43 8á80 16 6á80 1á78 5á67

The layer numbers correspond to the sample layers depicted in Figs 9 and 12, numbered consecutively from theconcretion top. Axial directions are relative to an arbitrary horizontal direction. Other data as in Table 2.

Fig. 11. Continued.



crystal at the site of nucleation, or produced bysuppression of some lattice orientations throughgrowth competition of adjacent crystals (Dickson,1993).

Fabrics which grow by syntaxial overgrowthson previous grains can only modify the aggregatePCO by a restriction of grain growth of adjacentgrains (with different PCO), as a result of compe-

tition or physical barriers. Preferential crystallo-graphic growth mechanisms could potentiallymodify an initial PCO if the crystals were com-peting for space on preferentially oriented sub-strates, and there was a preference for growth of aparticular oriented lattice (Dickson, 1993).Growth competition would only occur oncesuf®cient pore space had been ®lled by the

Fig. 12. Variation of AMS parame-ters h% and q with location inCrosby Warren Concretion F6. Eachbox represents the sample location.The shaded portion inside each boxrepresents either the h% or q-value,on a full scale of 0±20% for h% and0á0±2á0 for q. The arrows markedA-A¢ and B-B¢ indicate the positionof the vertical and horizontal geo-chemical spot traverses, which arerepresented in Fig. 10. Concretioncentre is at the mid, left-hand side.

550 M. W. Hounslow


growing crystallites, by which time any preferredorientation will have been dictated by the earliestgrowth phases. A possible mechanism respon-sible for preferential directional growth is `sidepoisoning', caused by various solute agents(Dickson, 1978; Lahann, 1978; Given & Wilkin-son, 1985). Such a growth-competition mechan-ism is unlikely in the concretions, because graincentres (initial crystallization centres?) appear tobe well-spaced apart and, therefore, initial com-petition for siderite growth space did not occur.Also, later growth appears to have encloseddetrital particles, which evidently did not act as

barriers to growth. In addition, such preferentialgrowth would probably produce elongate grainshapes, which are only present in the Hepworthconcretions. Consequently, this indicates theconditions during nucleation of the crystalliteswere fundamental in determining the resultingPCO of the fully grown grain and that of the ®nalaggregate.

Mechanism for producing preferredcrystallographic orientation

Two processes may have in¯uenced the nuclea-tion and growth of oriented siderite crystallitesduring formation of these concretions, stress-induced anisotropy and substrate-induced aniso-tropy.

Stress-induced anisotropy

The thermodynamic theory of crystallization in astress ®eld has been developed by Kamb (1959)and more recently by Patterson (1973) andMcKenzie et al. (1996). This has found applica-bility in predicting the preferred crystallographicorientation of carbonates and quartz recrystalli-sing under wet conditions in a stress ®eld, both inthe laboratory and in rocks (Neumann, 1969;Tullis & Yund, 1982; Ishill, 1988). The stress hastwo consequences for a growing crystal: (a) itindicates which crystallographic orientation ismost stable with reference to the deviatoric stress;and, (b) there is a chemical potential set-up foranisotropic growth of the crystal. The most stablecrystal orientation under a deviatoric stress ®eldis dependent on the crystal elastic constants, andis independent of hydrostatic pressure (Patterson,1973). Kamb (1959) predicts calcite to orient withthe c-axis parallel to the maximum stress, withgrowth along the minimum stress direction pre-ferred over that parallel to the maximum stress.Other carbonates belonging to the same crystalsystem should also show similar responses tostress, because the crystal elastic constants willbe broadly similar (Patterson, 1973; McKenzieet al., 1996).

During the conditions of normal sedimentconsolidation, the maximum stress is verticaland the horizontal stresses are typically 0á45±0á7of the effective overburden stress (Hounslow,1997). These features of the stress ®eld suggestthat cements formed within a clastic framework,which applies stress to the growing grains, wouldproduce a fabric with vertical carbonate c-axes.This explanation was utilized by Attewell et al.

Fig. 13. (A) Typical variation of magnetic susceptibil-ity (K) between liquid nitrogen and room temperaturefor sample from concretion IS19. The departure fromlinearity of the 1/K vs. Temperature at �100°K is due tothermal gradient effects across the specimen. (B) Vari-ation of the magnetic high ®eld torque with the squareof the applied ®eld (H2). The slight departure fromtorque-H2 linearity is probably due to changes in mag-netic ®eld homogeneity with H.



(1969) to explain X-ray texture determined dolo-mite and siderite fabrics with vertical c-axes.

Substrate-induced anisotropy

The mechanics of crystal nucleation are complexbut better understood in simple laboratory sys-tems than in natural systems (Nancollas & Purdie,1964; Sunagawa, 1994). In simple systems, mi-crotopological features of the nucleation substrateand the energetics of the solute±solvent andsubstrate interaction are important in controllingcrystal growth (Lagnally, 1993). In systems wherethe substrate is not the same as the crystallite,nucleation is dictated by two factors (Archibaldet al., 1996); (a) the ability of the substrate tomimic the structure of a matching crystallite face,

and (b) the stereochemistry and orientation ofmolecules at the interface. Work on biomineral-ization has shown that organic monolayers canprovide diverse substrates for carbonate nuclea-tion (Belcher et al., 1996; Aizenberg et al., 1999).Using various organic molecules as substrateconditioners, Archibald et al. (1996) demonstra-ted that calcite crystallized with either of thefollowing crytallographic planes parallel to thesubstrate; (a) {001} or planes near the {001} axis ofthe calcite (b) {110},{113} and {116}, which areplanes at high angles to the c-axis; (c) {104} ±c-axis 44á5° to the substrate. The propagation ofsuch fabrics in siderite nucleated on a substratewould produce for the three states above; (a) aprolate fabric with the Kmax, axis perpendicular tothe substrate; (b) a subhorizontal oblate fabric

Fig. 14. Schematic of the suggested c-axis pole density distributions for siderite, represented on stereographicprojections (darkest ornament is highest concentration of c-axes). The interpreted average crystal orientations areshown relative to horizontal and the resulting position of the AMS axes. (A) The pole-density distribution for thevertical Kmax axis fabric, produced by c-axes vertical to the bedding (Crosby Warren concretions). The eccentricity inthe distribution would give rise to the anisotropy within the bedding plane. (B) C-axes distributed about a planeparallel to the bedding, produces Kmin vertical and Kmax within the bedding plane (Hepworth concretions). (C) Girdledistribution of c-axes, which could give rise to a nearly isotropic, or weak triaxial AMS fabric. (D) Form of subfabricsuggested for centre of F6, which is similar to C but with an extra horizontal biaxial concentration of c-axes.

552 M. W. Hounslow


with Kmin perpendicular to the substrate; and (c)no AMS or a weak triaxial fabric.

Clearly, the nucleation and substrate interac-tion process has the potential to modify thelattice orientation of a crystallite forming on asubstrate. A substrate control for dolomite hasalso been demonstrated by Toman & Taylor(1974), which oriented with the c-axis normalto the substrate. The nature of the substrate alsocontrolled the Mg content of the dolomite. Thesedemonstrate that substrate controls are likely toin¯uence the nucleation of siderite. In addition,difference between fresh and saltwater pore ¯uidchemistry and the changing (diagenesis medi-ated) chemical balance of the pore water systemscould have the potential to produce a variety ofcrystalline fabrics. The action of Mg, SO4

± andPO4

±, often quoted as `surface poisons' (Lahann,1978; Given & Wilkinson, 1985), also has theability to modify the crystallite morphology(Paquette et al., 1996).

Substrate vs. non-substrate controllednucleation and growth

Attewell et al. (1969) and Bellamy (1980) haveexplained the origin of the vertical c-axis fabric ofcarbonates in marine sediments and concretionsas having formed in response to a compactioninduced stress ®eld. This explanation predicts anincrease in the degree of PCO of the fabric acrossconcretions, corresponding to an increase information depth, concomitant with an increasein the differential stress, leading to an increase inthe differential chemical potentials which drivethe orienting mechanism. However, there are anumber of reasons why this mechanism is notlikely to be important:

1 Applying the same reasoning to the Hep-worth concretions would indicate the maximumdeviatoric stress was horizontal during theirformation. Horizontal maximum stress is incom-patible with predominantly extensional syn-tectonics during the late Carboniferous (Guion &Fielding, 1988; Lee, 1988). Horizontal maximumstress is feasible under conditions of erosionalunloading, producing overconsolidated sedi-ments (Brooker & Ireland, 1965; Hounslow,1997). The horizontal maximum deviatoric stressto produce the horizontal c-axis fabric in theHepworth concretions, would also need increas-ing differential stress over the growth period ofthe concretion, to produce the centre-peripheryincrease in the degree of orientation. However,

conditions of horizontal maximum stress, if theyexisted, would probably be transient (Mayne &Kulhawy, 1982), and incompatible with a pro-gressive growth and burial model for the Hep-worth concretions.

2 The most favourable site for nucleation willbe a substrate. Consequently, siderite crystalliteswhen initially formed on a substrate are likely tobe surrounded by pore water on three sides, andtherefore would experience hydrostatic pressurerather than the deviatoric stress ®eld transmittedthrough the grain framework. Later siderite graingrowth appears to have been syntaxial over-growths, so the initial crystallite orientation andsubstrate conditioning is crucial in determiningthe initial PCO.

3 The chemical potentials driving the orientingmechanism involved in Kamb's theory, are verysmall (also noted by Etheridge et al., 1974), evenfor the large stresses involved in tectonic defor-mation. Therefore, the relatively small deviatoricstresses during early compaction (Hounslow,1997) are likely to be too small to producesigni®cant PCO.

4 The anisotropic growth rates suggested byKamb's theory indicate that grain shape would beelongate perpendicular to the c-axis, as opposedto the length-fast grains present in the Hepworthnon-marine concretions, and the equant grains inthe Crosby Warren marine concretions.

5 There is evidence from concretion F6 andother studies (Attewell et al., 1969; Bellamy,1980), that fabrics other than vertical c-axisfabrics are found in marine concretions.

Therefore, the siderite fabric of the concretionsappears to be determined by substrate surfaceproperties during nucleation of the siderite crys-tallites. In order to produce a PCO by nucleationon a substrate, the substrate itself must have apreferred orientation. In the case of naturalnucleation the substrates are most likely to beclay minerals (Berner, 1980; Morse, 1986), condi-tioned with the complex inorganic/organic soupin the pore space, consequently the siderite fabricwill re¯ect the preferred orientation of theseparticles.

To produce the Hepworth non-marine concre-tion fabrics, the substrate-siderite attachmentplane must be at a small angle to the c-axis(Fig. 14B). Because the clay minerals show agreater degree of preferred orientation towardsthe outside of these concretions (Oertel & Curtis,1972), the siderite crystals would mimic thisorientation by a greater degree of orientation. If



such a process is operative, then differencesbetween and within concretions should re¯ectdifferences in the preferred orientation of thesubstrate. Con®rmation of this is found by thefabric data for IS25 (cemented sandy siltstone),which has lower h% values (Table 2). Thiscorresponds to the lower degree of orientation ofthe clay fraction in the coarser grained lithologies,as measured by Curtis et al. (1980) in the Hep-worth sequence. The h% changes across IS25 mayre¯ect differences in the clastic grain size, andtheir effect on the orientation of the clay sub-strates (Curtis et al., 1980; Moon & Hurst, 1984).

Fisher et al. (1998) re-addressed the growthmodel for the Hepworth siderite concretions,suggesting that growth could be initiated andcontinue at the same depth within a methano-genic environment. In addition they suggested,that the clay fabric data of Oertel & Curtis (1972)shows the effects of differential compactioncaused by differences in the amount of cement-ing siderite, rather than the effect of differentburial depths. There are two problems with thisargument. First, the clay and siderite fabricsshow progressive, rather than abrupt changeswith similar siderite abundance through thecemented concretion bodies (e.g. Figs 3 and 4;Oertel & Curtis, 1972), suggesting burial andcementation were progressive. Second, thezoned siderites and compositionally differentsiderite crystallite centres (e.g. Fig. 2A and C)imply adjacent crystals formed at different times,according to the model of Fisher et al. (1998), orunder different bacterial activity rates (age ordepth?) according to the data of Mortimer et al.(1997). Hence, chemical and 13dC changes alonein these concretions can not be con®dently usedto indicate differences in burial depth. However,the combination of centre-periphery fabric chan-ges, with early (Mn-enriched) followed by later(Mg enriched) siderite, indicates a time differ-ence, which would in all probability be accom-panied by more compaction.

Fabrics as indicators of diagenetic stage

There is evidence that late and early formeddiagenetic marine carbonates have different fab-rics. Bellamy (1980), in a study of marine concre-tionary carbonates from the UK Jurassic, noted avariety of crystallographic and shape fabrics,using optical observations and a simple X-raymethod for determining approximate c-axis ori-entations. He found horizontal and almost ran-dom c-axis fabrics in very early formed calcite

and dolomite concretions. Marine dolomite andsiderite concretions formed later in diagenesispossessed a vertical c-axis fabric (Bellamy, 1977;Irwin et al., 1977). The X-ray fabric measure-ments of Attewell et al. (1969) on dolomiteshowed a similar story, with a dominant verticalc-axis fabric, and a subsidiary girdle grouping ofc-axes at �44° from the vertical. They suggestedthat early formed dolomite showed a preferencefor orienting with a rhomb face aligned parallel tothe bedding, such as that depicted in Fig. 14C.Fuchtbauer & Lindenburg (1969) have alsomeasured similar fabrics in calcite, which theyexplained as resulting from rhombohedral oracute rhombohedral crystal faces oriented paral-lel to the strati®cation.

It is proposed that the initial fabric produced inthe centre of concretion F6 is an inclined asym-metric girdle of c-axes. This could be produced byrhombohedral faces (1011 crystal form) of sideriteparallel to bedding, essentially similar to thatmeasured by Attewell et al. (1969) for earlydolomite (Fig. 14C). In order to explain the F6fabric an additional biaxial concentration(Fig. 14D) in the girdle or bedding plane wouldbe needed to give rise to the AMS in the beddingplane. The strength of the asymmetric girdledecreases in magnitude away from the concretioncentre, as indicated by the increase in the q-value(Fig. 12). In addition to this fabric, the latersiderite (which is also present in the concretioncentre) has a strong vertical c-axis orientation,such as that in concretion F5. This would implythe F5 fabric formed later than most of the sideritein F6, which is supported by the lower sideritecontents in F5.

Bringing together these fabric elements it ispossible to qualitatively explain the AMS featuresmeasured in concretion F6. This interpretationinvolves three distinct fabric types, the spatialcombination of which is responsible for thevariation measured.

(a) Vertical c-axes of siderite concentratedmainly in the outer half of the concretion, butpresent throughout the concretion body, which isperhaps the last formed fabric type (Fig. 14A).

(b) A symmetric girdle of c-axes oriented about44° from the vertical, which is most concentratedin the region of low h% values around theconcretion centre (e.g. Figure 14C). As a resultof the axial representation used by AMS thisfabric is seen as a poor degree of preferredorientation, hence the lower h% values in thisregion.

554 M. W. Hounslow


(c) A horizontal biaxial or inclined biaxialfabric (Fig. 14D) which is present in the concre-tion centre and appears to be closely associatedwith the girdle fabric (b).

To generate the measured AMS there has to beconsiderable intergrowth of these three fabrics.The complex variation of the AMS parametersh% and q indicates the three fabrics are inter-mixed, perhaps related to the two siderite phasespresent in this concretion.

The origin of the biaxial concentration in F6responsible for the AMS in the bedding plane isnot clear. It may be related to either a inherentclastic particle anisotropy, which affects thesubstrate orientation, such as the AMS fabricmodel based on compaction, proposed byHounslow (1986) for a chlorite fabric. Alternat-ively it may be controlled by oriented condition-ing molecules on the nucleation substrate (seeArchibald et al., 1996). There is some supportfor the former likelihood, in that the Hepworthconcretions (HA and HA-1) show mean Kmax

oriented approximately E-W, which is coinci-dent with delta progradation and ¯uvial channelsystems in the Westphalian A (Guion & Fielding,1988). Such an orientation of Kmax axes would beproduced if siderite nucleated on a compaction-induced clay fabric, impressed onto a detritalgrain orientation such as the model proposed byHounslow (1986).

Wider implications for carbonate fabrics

An interpretation based on substrate control ofthe fabric provides a single mechanism wherebythe fabrics of the marine and non-marine sideritescan be understood. This interpretation would alsounify the dominant presence of c-axis verticalfabrics in marine concretions, with the fact thatmarine carbonate rim cements in limestonescommonly have c-axes oriented normal to thesubstrate (Dickson, 1978, 1993); although thereare exceptions to this observation (Lindholm,1974). If such a comparison is warranted itimplies that the mechanics that control theorienting mechanism for carbonates are largelyindependent of the rock type in which they form;rather they may characterize particular surfaceproperties of natural particles and the formingcrystallite. A surface-reaction control on thegrowth of calcite concretions as proposed byRaiswell (1988), is entirely compatible with thenucleation-substrate conditioning proposed forthe generation of these siderite fabrics.

CONCLUSIONS

The concretions show evidence for dual sideritephases, which in the non-marine concretions arealso evident as zoned siderite. Compositionaldifferences are also shown by adjacent sideritegrain centres, indicating a probable difference intime of crystal formation. Such siderite texturaldifferences are indicative of the concretion ini-tially forming as a diffuse patch of carbonatecement, but through time (and greater burialdepths) the concretion gradually developed byovergrowths on the initial crystallites, and addi-tion of newly nucleated crystals on adjacentsubstrates.

The variety of fabric types in the marine andnon-marine concretions are best explained bychemical conditioning of the nucleation substraterather than stress induced anisotropy. The form ofthe siderite crystallographic fabric is probably afunction of both the timing of diagenesis andnucleation substrate conditioning. The differingchemistry and depth/time evolution of marineand non-marine pore waters could explain thebasis of the different fabrics from these twosettings, where the nucleation substrate condi-tioning determines the orientation behaviour ofthe siderite. The conditioning determines theorientation of the nucleating crystallites on thesubstrate. The preferred orientation of the nucle-ation substrate ultimately dictates the resultingpreferred crystallographic orientation of the car-bonate species.

ACKNOWLEDGEMENTS

Charles Curtis allowed access to his Westphalianconcretion collection. This manuscript bene®tedfrom discussions with Phil Newton and thecomments of Alan Kendall, Peter Mozley, SadoonMorad and Tim Astin. The comments of Sedi-mentology reviewers, Keith Benn, Ray Ferrell,and Tim White also improved the manuscript.Hepworth Minerals allowed access to their quar-ries. The high ®eld torque measurements wereperformed at Southampton University, Oceanog-raphy Department.

REFERENCES

Aizenberg, J., Black, A.J. and Whitesides, G.M. (1999) Control

of crystal nucleation by patterned self-assembled monolay-

ers. Nature, 398, 495±498.



Allison, P.A. and Pye, K. (1994) Early diagenetic mineralisa-

tion and fossil preservation in modern marine carbonate

concretions. Palaios, 9, 561±575.

Archibald, D.D., Qadri, S.B. and Gaber, N.P. (1996) Modi®edcalcite deposition due to ultrathin organic ®lms on silicon

substrates. Langmuir, 12, 538±546.

Ashby, D.A. and Pearson, M.J. (1979) Mineral distribution in

sediments associated with the Alton Marine Band nearPenistone, South Yorkshire. Dev. Sedimentol., 27, 311±321.

Attewell, P.B., Hirst, D.M. and Taylor, R.K. (1969) Diagenetic

recrystallisation and orientation of two carbonate species.Sedimentology, 14, 237±247.

Belcher, A.M., Wu, X.H., Christensen, R.J., Hansma, P.K.,Stucky, G.D. and Morse, D.E. (1996) Control of crystal phase

switching and orientation by soluble Mollusc-shell pro-teins. Nature, 381, 56±58.

Bellamy, J.R.W. (1977) Surface expansion megapolygons in

Upper Jurassic dolostone (Kimmeridge, UK). J. Sed. Petrol.,47, 973±978.

Bellamy, J.R.W. (1980) Carbonates within bituminous shalesof the British Jurassic: their petrography and diagenesis.Unpubl. PhD Thesis, University of Southampton, UK.

Berner, R.A. (1980) Early Diagenesis, a Theoretical Approach.

Princeton University Press, Princeton.

Brooker, E.W. and Ireland, H.O. (1965) Earth pressures at rest

related to stress history. Can. Geotech. J., 2, 1±15.Coleman, M.L. (1985) Geochemistry of diagenetic non-silicate

minerals: kinetic considerations. Phil. Trans. Roy. Soc.London, 315A, 39±56.

Coleman, M.L. and Raiswell, R. (1995) Source of carbonateand origin of zonation in pyritiferous carbonate concretions-

evaluation of a dynamic model. Am. J. Sci., 295, 282±308.

Curtis, C.D., Pearson, M.J. and Somogyi, V.A. (1975) Miner-alogy, chemistry and origin of a concretionary siderite sheet

(clay-ironstone band) in the Westphalian of Yorkshire.Mineral. Mag., 40, 385±393.

Curtis, C.D., Lipshie, S.R., Oertel, G. and Pearson, M.J. (1980)Clay orientation in some Upper Carboniferous mudrocks, its

relationship to quartz content and some inferences about

®ssility, porosity and compactional history. Sedimentology,

27, 333±339.Curtis, C.D., Coleman, M.L. and Love, L.G. (1986) Pore water

evolution during sediment burial from isotopic and mineral

chemistry of calcite, dolomite and siderite concretions.

Geochim. Cosmochim. Acta, 50, 2321±2334.Dickson, J.A.D. (1978) Length slow and length-fast calcite:

a tale of two elongations. Geology, 6, 560±561.

Dickson, J.A.D. (1993) Crystal-growth diagrams as an aid tointerpreting the fabrics of calcite aggregates. J. Sed. Petrol.,63, 1±17.

Etheridge, M.A., Paterson, M.S. and Hobbs, B.E. (1974)

Experimentally produced preferred orientation in synth-etic mica aggregates. Contrib. Mineral. Petrol., 44, 275±294.

Fisher, Q.J., Raiswell, R. and Marshall, J.D. (1998) Siderite

concretions from non-marine shales (Westphalian A) of the

Pennines, England: Controls on their growth and composi-tion. J. Sed. Res., 68, 1034±1045.

France, D., Hu, Y., Snowball, I., Rolph, T., Old®eld, F. and

Walden, J. (1999) Additional rock magnetic measurements.In: Environmental Magnetism: a Practical Guide (Eds J.

Walden, F. Old®eld and J.P. Smith), Quatern. Res. Assoc.,Technical Guide, 6, 197±211.

Fuctbauer, H. and Lindenburg, H.-G. (1969) Lower Carboni-ferous `Kulm Plattenkalk' between Reh and Muschede,

Westfalen, West Germany. In: Carbonate Diagenesis (Eds

F.T. Mackenzie, R.N. Ginsburg, L.S. Land and O.P. Bricker),

Bermuda Biological Station for Research Spec. Publ., 3,128±131.

Gaunt, G.D., Fletcher, T.P. and Wood, C.J. (1992) Geology ofthe Country Around Kingston Upon Hull and Brigg. British

Geological Survey memoir, for sheets 80 and 89. HMSO,

London.

Gautier, D.L. (1982) Siderite concretions: indicators of earlydiagenesis in the Gammon Shale (Cretaceous). J. Sed. Pet-rol., 52, 859±871.

Given, R.K. and Wilkinson, B.H. (1985) Kinetic control ofmorphology, composition and mineralogy of abiotic sedi-

mentary carbonates. J. Sed. Petrol., 55, 109±119.

Guion, P.D. and Fielding, C.R. (1988) Westphalina A and B

sedimentation in the Pennine Basin, UK. In: Sedimentationin a Synorogenic Basin Complex: the Upper Carboniferousof Northwest Europe (Eds B.M. Besley and G. Kelling),

pp. 153±177. Blackie and Son, Glasgow.

Hounslow, M.W. (1984) Sedimentological implications ofmagnetic fabrics measurements of Jurassic and late Triassicsediments, UK. Unpubl. PhD Thesis, University of Sou-

thampton, UK.Hounslow, M.W. (1986) Magnetic fabric arising from para-

magnetic phyllosilicate minerals in mudrocks. J. Geol.Soc. London, 142, 995±1006.

Hounslow, M.W. (1997) Signi®cance of localised pore pres-sures to the genesis of septarian concretions. Sediment-ology, 44, 1133±1147.

Hounslow, M.W., Noel, M. and Bootes, P.A. (1988) Sensitivity

and sample shape related measuring effects on the Molspinsusceptibility anisotropy meter. Geophys. J. Roy. Astron.Soc., 94, 355±363.

Hrouda, F. (1980) Magnetocrystalline anisotropy of rocks andmassive ores: a mathematical model study and its fabric

implications. J. Struct. Geol., 2, 459±462.

Irwin, H., Curtis, C.D. and Coleman, M. (1977) Isotopic evi-

dence for source of diagenetic carbonates formed duringburial of organic-rich sediments. Nature, 269, 209±213.

Ishill, K. (1988) Grain growth and reorientation of phyllosili-

cate minerals during the development of slaty cleavage in

the South Kitakami Mountains, N.E. Japan. J. Struct. Geol.,10, 145±154.

Jacobs, I.S. (1963) Metamagnetism of siderite (FeCO3). J. Appl.Phys., 34, 1106±1107.

Jones, B. and Kahle, C.F. (1993) Morphology, relationship andorigin of ®ber and dendrite calcite crystals. J. Sed. Petrol.,63, 1018±1031.

Kamb, W.B. (1959) Theory of preferred crystallographic ori-entation developed under stress. J. Geol., 67, 153±170.

Kristman, K.S., Chakravorty, N.C. and Banerjee, S. (1932)

Investigation in magneto-crystallic action, Part II, paramag-

netics. Phil. Trans. Roy. Soc. London, 232A, 99±115.Lagnally, M.G. (1993) Atom motions on surfaces. Phys. Today,

46, 24±31.

Lahann, R.W. (1978) A chemical model for calcite crystal

growth and morphology control. J. Sed. Petrol., 48, 337±344.Lee, A.G. (1988) Carboniferous basin con®guration of

central and northern England modelled using gravity data.

In: Sedimentation in a Synorogenic Basin Complex:the Upper Carboniferous of Northwest Europe (Eds B.M.

Besley and G. Kelling), pp. 69±84. Blackie and Son,

Glasgow.

Lindholm, R.C. (1974) Fabric and chemistry of pore-®llingcalcite in septarian veins: models for limestone cementa-

tion. J. Sed. Petrol., 44, 428±440.

556 M. W. Hounslow


Longman, M.W. (1980) Carbonate diagenetic textures from

near surface diagenetic environments. AAPG Bull., 64,461±487.

Love, L.G., Coleman, M.L. and Curtis, C.D. (1983) Diageneticpyrite formation and sulphur isotope fractionation associ-

ated with a Westphalian marine incursion, Northern

England. Trans. Roy. Soc. Edinburgh, 74, 165±182.

Mardia, K.V. (1972) Statistics of Directional Data. AcademicPress, New York.

Matsumoto, R. and Iijima, A. (1981) Origin and diagenetic

evolution of Ca-Mg-Fe carbonates in some coal®elds ofJapan. Sedimentology, 28, 239±259.

Mayne, P.W. and Kulhawy, F.H. (1982) K0-OCR relationships

in soil. ASCE J. Geotech. Eng. Division, 108, 357±366.

McKenzie, D.R., Yin, Y., McFall, W.D. and Hoang, N.H.(1996) The orientation dependence of elastic strain energy

in cubic crystals and its implication for the preferred ori-

entation in titanium nitride. J. Phys. ± Condensed Matter, 8,5883±5890.

Moon, C.F. and Hurst, C.W. (1984) Fabric of mud and shales:

an overview. In: Fine Grained Sediments: Deep Water Pro-cesses and Facies (Eds D.A.V. Stow and D.J.W. Piper), Geol.Soc. London Spec. Publ., 15, 579±594.

Morse, J.W. (1986) The surface chemistry of calcium carbonate

minerals in natural waters: an overview. Mar. Chem., 20,91±112.

Mortimer, R.J.G., Coleman, M.L. and Rae, J.E. (1997) Effect of

bacteria on the elemental composition of early diagenetic

siderite: implications for palaeoenvironemntal interpret-

ations. Sedimentology, 44, 759±765.Mozley, P.S. (1996) The internal structure of carbonate con-

cretions in mudrocks: a critical evaluation of the conven-

tional concentric model of concretion growth, Sed. Geol.,103, 85±91.

Mozley, P.S. (1989a) Complex compositional zonation in

concretionary siderite: implications for geochemical stud-

ies. J. Sed. Petrol., 59, 815±818.Mozley, P.S. (1989b) Relation between depositional environ-

ment and the elemental composition of early diagenetic

siderite. Geology, 17, 704±706.

Mozley, P.S. and Burns, S.J. (1993) Oxygen and carbon iso-topic composition of marine carbonate concretions ± an

overview. J. Sed. Petrol., 63, 73±83.

Nancollas, G.H. and Purdie, N. (1964) The kinetics of crystal

growth. Quatern. Sci. Rev., 18, 1±20.Neumann, E. (1969) Experimental recrystallisation of dolomite

and comparison of preferred orientations of calcite and

dolomite in deformed rocks. J. Geol., 77, 426±438.Oertel, G. and Curtis, C.D. (1972) Clay ironstone concretion

preserving fabrics due to progressive compaction. Geol. Soc.Am. Bull., 83, 2597±2606.

Owens, W.H. and Bamford, D. (1976) Magnetic, seismic andother anisotropic properties of rock fabrics. Phil. Trans. Roy.Soc. London, 283A, 55±68.

Paquette, J., Vali, H. and Mucci, A. (1996) TEM study of

Part-C replicas of calcite overgrowths precipitated fromelectrolyte solutions. Geochim. Cosmochim. Acta, 60,4689±4699.

Patterson, M.S. (1973) Nonhydrostatic thermodynamics andits geologic application. Rev. Geophys. Space Phys., 11,355±389.

Pearson, M.J. (1974) Magnesian siderite in carbonate concre-

tions from argillaceous sediments in the Westphalian of

Yorkshire. Mineral. Mag., 39, 700±704.

Pearson, M.J. (1979) Geochemistry of the Hepworth Carbon-iferous sediment sequence and the origin of the diagenetic

iron minerals and concretions. Geochim. Cosmochim. Acta,

43, 927±941.

Pearson, M.J. (1985) Some chemical aspects of diageneticconcretions from the Westphalian of Yorkshire, England.

Chem. Geol., 48, 231±241.

Pedley, M., Andrews, J., Ordonez, S., delCura, M.A.G., Mar-tin, J.A.G. and Taylor, D. (1996) Does climate control the

morphological fabric of freshwater carbonates? A compar-

itive study of Holocene barrage tufas from Spain and Brit-

ain. Palaeogeogr. Palaeoclimatol. Palaeoecol., 121, 239±257.Powell, H.E. and Miller, C.K. (1962) Magnetic susceptibility of

siderite. US Bureau Mines Report Invest., 6224, 8.

Pye, K. (1984) Rapid estimation of porosity and mineral

abundance in backscattered electron images using a simpleSEM image analyser. Geol. Mag., 121, 81±136.

Raiswell, R. (1988) Evidence for surface reaction-controlled

growth of concretions in shales. Sedimentology, 35, 571±575.Reed, S.J.B. (1996) Electron Microprobe Analysis and Scan-

ning Electron Microscopy in Geology. Cambridge University

Press, Cambridge.

Richter, C. and Van der Pluijm, B.A. (1994) Separation ofparamagnetic and ferrimagnetic susceptibility using low

temperature magnetic susceptibility and comparison with

high ®eld methods. Phys. Earth Planet. In., 82, 113±127.

Richter, C., Van der Pluijm, B.A. and Housen, B.A. (1993) Thequanti®cation of crystallographic preferred orientation

using magnetic anisotropy. J. Struct. Geol., 15, 113±116.

Rochette, P. (1987) Magnetic susceptibility of the rock matrixrelated to magnetic fabric studies. J. Struct. Geol., 9, 1015±

1020.

Rochette, P. (1988) Inverse magnetic fabric in carbonate bear-

ing rocks. Earth Planet. Sci. Lett., 90, 229±237.Rochette, P., Jackson, M. and Auboug, C. (1992) Rock mag-

netism and the interpretation of anisotropy of magnetic

susceptibility. Rev. Geophysics, 30, 209±226.

Spears, D.A. (1989) Aspects of iron incorporation into sedi-ments with special reference to Yorkshire ironstones. In:

Phanerozoic Ironstones (Eds T.P. Young and W.E.G.

Taylor), Geol. Soc. London Spec. Publ., 46, 60±71.

Sunagawa, I. (1994) Nucleation, growth and dissolution ofcrystals during sedimentogenesis and diagenesis. In: Dia-genesis IV, Chapter 2 (Eds K.U. Wolf and G.V. Chilingarian),

Dev. Sedimentol., 51, 19±47.Tarling, D.H. and Hrouda, F. (1993) The Magnetic Anisotropy

of Rocks. Chapman & Hall, London.

Taylor, K.G. and Curtis, C.D. (1995) Stability and facies

association of early diagenetic mineral assemblages ± anexample from the Jurassic ironstone-mudstone successsion,

U.K. J. Sed. Res., 65, 358±368.

Toman, K. and Taylor, P.M. (1974) Epitaxial growth of dolo-

mite on mica. Am. Mineral., 59, 871±872.Tullis, J. and Yund, R.A. (1982) Grain growth kinetics of quartz

and calcite aggregates. J. Geol., 90, 308±318.

Manuscript received 10 November 1999;revision accepted 16 August 2000.



The crystallographic fabric and texture of siderite in concretions: implications for siderite nucleation and growth processes

Documents