Sei sulla pagina 1di 20

Sedimentology (1980) 27, 63 1-650

Stages in a Precambrian dolomitization, Scotland: cementing versus


replacement textures

I A N J. F A I R C H I L D *
Department of Geology, University Park, Nottingham NG7 2RD, and
Department of Geology, Downing Street, Cambridge CB2 3EQ

ABSTRACT

Dolomitization in member 1 of the Dalradian Bonahaven Formation of lslay was sequential. The
sediments studied are siliciclastic tidal flat deposits, which originally contained pebbles of ferroan
dolomicrite and sandy dolomicrite (type 1 dolomite). Following burial, both dolomicrite and clay-
stone pebbles, and possibly other carbonates, were replaced by coarser, fluid-inclusion-bearing,
chemically zoned (type 2) dolomite. Ferroan, inclusion-free, intergranular dolomite (type 3) in
sandstones may be synchronous with late zones of type 2 dolomite which it closely resembles.
Cathodoluminescence and microprobe analyses of type 2 dolomite characterize four ubiquitous
growth stages within which there was considerable fluctuation in Fe and Mn content. There was much
variation between and within samples in terms of timing of nucleation, absolute speed of growth and
relative growth rate of crystals. Type 2 dolomite displays two types of intracrystalline truncation
surfaces explicable by dissolution followed by overgrowth, or by replacement by a moving solution
film. Type A is characterized by a single later zone cutting across earlier ones along the line of an
irregular truncation surface which developed from the edges of the crystals. In type B truncations it is
the interior, rather than the edges of crystals which show truncation, the truncations forming smooth
closed surfaces which enclose areas filled centripetally with new zones. In general, dolomite has a
pronounced tendency to form crystal faces, and thus compromise crystal boundaries, during replace-
ment as well as during cementation. Thus, if nucleation of crystals is restricted to the margins of an
area to be replaced, replacive dolomite is capab!e of completely mimicking a vug-filling texture.

INTRODUCTION

This paper examines a case of diagenetic dolomitiz- of crystal nuclei, and the variable speed of crystal
ation which seems atypical in that it occurs within a growth, can thus be appreciated. This leads t o a
siliciclastic sequence and does not fit any current discussion of the possibility that replacive dolomite,
large-scale dolomitization model. Its wider sig- with its pronounced ability to form crystal faces, can
nificance relates to features on the microscopic scale in some circumstances mimic cement textures. I n
illustrating the intricate history of progressive addition, two types of intracrystalline truncation
dolomitization. This paper is one of the first studies texture are recognized, one of which broadly
to utilize a combination of cathodoluminescence resembles the truncations caused by corrosion and
petrography and microprobe analysis in a diagenetic overgrowth in cements recorded by Meyers (1978),
problem. The complex nature of dolomitization o n a the other seems completely new.
small scale in terms of timing and spatial distribution The dolomitized strata lie a t the top of member 1
* Present address : Department of Geology, University of the Bonahaven Formation (Fig. 1; Spencer &
of Birmingham, P.O. Box 363, Birmingham B15 2TT. Spencer, 1972) which belongs t o the Dalradian
0037-0746/80/1200-0631 $02.00 Supergroup and is probably of latest Precambrian
0 1980 International Association of Sedimentologists (Vendian) age (Harris & Pitcher, 1975). The rocks
632 I. J. Fairchild

have suffered chlorite-grade regional metamorphism, Microprobe analysis was carried out on gold-
but the particular horizon described here shows no coated polished thin sections on a Cambridge
trace of deformation and metamorphism, except Microscan V housed in the Wolfson Institute of
within some claystone pebbles (see caption to Interfacial Technology at Nottingham University.
Fig. 5d). As carbonates decompose under the electron beam
The member 1 dolomite forms pebbles and inter- they present special problems in quantitative analy-
granular material and is mostly coarsely crystalline. sis, which become more acute when analysing finely
It forms a stratigraphically distinct occurrence from zoned crystals. The method employed in this work
the highly dolomitic member 3 where the dolomite is makes use of continuous scans and a high specimen
overwhelmingly fine-grained (Fairchild, 1980). current under which the specimen visibly decom-
poses. However, by using carbonate standards of
similar minor element content to that of the speci-
METHODS men, both specimen and standard decompose to a
similar degree and the necessity for a ZAF correc-
About thirty samples were studied in thin section, tion procedure is eliminated. The scan lines are
about half of these using cathodoluminescence visible optically and allow a ready comparison of the
petrography. Thin sections from four of the speci- microprobe results with the zonal stratigraphy
mens (all from the top 5 rn of member 1 in Section B) revealed by luminescence. The detection limits are
were analysed by an electron microprobe prior to also very low (calculated by the method of Reed,
luminescence. 1973, to be about 20 p.p.m. for Fe and Mn in

J u r a Ouortzite
Formation

Port Askaig
T i l l i t e Formation

sandstone

.,....
..... intruformational conglomerate

44 snndstone with slit a n d mud


.....
.....
...:.
11 i n t e r b ed s ( 3 - 50c m thic k l

occurrence of type 1 & 2


dolomite

?
occurrence of type 3
6 dolomite in sandstones
- ubiquitous
- patchy
r7
V
SECTION A SECTION B
Fig. 1. General lithology and distribution of dolomite in the coastal sections of member 1 of the Bonahaven Formation.
Scales of sections indicate metres above the base of the Bonahaven Formation. National Grid references of tops of
sections: A, NR 4290 7040; B, NR 4235 7284.
Dolomite textures 633

dolomite). Operating conditions : accelerating volt- ments, or highly dolomitic sandstones. The upper
age 15-16 kV, specimen current 380 nA, beam size limit of dolomite occurrence, unlike the lower is
4 4 . 5 pm (focused), specimen scan mode (usually at rather sharp, but does not appear to mark a change
10 pm min-l). Totals deviating from 100% were in environmental conditions since it occurs within an
attributed to differential decomposition of standard otherwise uniform sequence of cross-stratified sand-
and specimen and were recalculated to 100% as was stones.
done by Moberly (1968, 1970). Member 1 wedges out to the west between
Cathodoluminescence petrography was carried siliceous sandstones above and pebbly sandstones
out on home-made instruments at the Universities and mixtites below and is not present further than
of Nottingham and Cambridge. Experimental details 5 km inland; dolomitic sediments occur at the top of
for photomicrography (after Marshall, 1978): 30 the member wherever it is seen.
kV; 8 mA; 1 cm diameter beam; twin cold-cathode
guns; hydrogen atmosphere; l o x ocular; 2.5 or
6.3 x objective; 35 pm thick polished thin sections;
Zeiss photomicroscope; exposure times typically GENERAL L I T H O L O G Y AND
6-9 min on 400 ASA film, developed normally. The PETROGRAPHY
presence or absence of luminescence is determined
by the chemical composition of the carbonate, being Dolomite occurs as clasts and intergranular material.
activated by Mn (Sippel & Glover, 1965) and There are three petrographic varieties of dolomite:
quenched by Fe (Sommer, 1972; Pierson, 1977). In fine-grained (5-15 pm), type 1; inclusion-bearing,
the suite of dolomites described here, 1 wt% Fe is chemically zoned, 15-300 pm, type 2 ; inclusion-free,
sufficient to damp out luminescence and about 30-300 pm, type 3. Their properties are summarized
0.02 wt% Mn is required to activate luminescence at in Table 1 and their stratigraphic distribution indi-
lower Fe concentrations. cated on Fig. 1 .
Claystone, as layers 1-50 mm thick, is the only
lithology lacking dolomite. Claystones consist largely
of phengite (illitic mica) together with detrital quartz,
STRATIGRAPHIC SETTING microcline and magnetite. Those claystones near the
top of the member (top 4 m in Section B) contain
Member 1 of the Bonahaven Formation is restricted 50-150 pm spherules of phengite interpreted as
to the Isle of Islay and is best exposed in two sections glauconitized microfossils (Fairchild, 1977). In
on the east coast (Fig. 1). The lithologies are contrast, claystone pebbles do sometimes contain
dominantly siliciclastic: towards the top of the dolomite crystals (type 2). Claystone pebbles are
member, parallel-laminated siltstones and mud- often rather deformed by compaction (Fig. 2a).
cracked claystones pass upwards into sandstones Siltstones occur interlaminated on a mm-scale
showing small-scale cross-stratification, siltstones, with claystones and as 1-6 cm beds grading up to
claystones, and conglomerates of intraformational claystone. Sandstone beds vary from very fine- to
pebbles. This is interpreted as a back-barrier coarse-grained and are 1-200 cm thick. Both silt-
sequence, washover fan and pond deposits passing stones and sandstones consist of clasts of quartz,
up into tidal flat sediments (Fairchild, 1977, 1978). microcline and subordinate magnetite. Intergranular
The succeeding member 2 consists of siliceous material is mica (representing squashed clay clasts),
sandstones, mostly parallel-laminated, but cross- or type 2 or 3 dolomite.
stratified at the base of the member. Member 2 is Conglomerates ( 3 4 0 cm thick) contain various
regarded as a transgressive sub-tidal deposit (Fair- intraformational pebbles (Fig. 2) : claystone, silty
child, 1978). No important breaks in sedimentation claystone, and dolostone (consisting of type 1 and/or
can be recognized within member 1 or above it for type 2 dolomite with or without sand or silt im-
at least several hundred metres. purities). The conglomerate matrix is dolomitic sand
Dolomite is present in clasts and intergranular like that of interbedded sandstones.
material in the upper part of member 1 (Fig. 1). Pure Peculiar features of sandy dolostone pebbles (Figs
dolostone beds appear to be present in the field, but 2b and 3c) and the sandy matrix of conglomerates
invariably prove on laboratory examination to be (Fig. 3e) are discrete areas of sparry dolomite.
tightly packed conglomerates of dolostone frag- Crystal growth zones (see later) always indicate an
634 I. J . Fairchild

Fig. 2. Thin sections of conglomerates showing most of the pebble types. c, claystone; SIC, silty claystone; 2, type 2
dolomite; sd, sandy dolostone (overwhelmingly type 1 dolomite in (a), mixture of type 1 and type 2 dolomite in (b) where
arrows also point to pseudo-vug-fills). The sd pebble in (a) is outlined for clarity. Such diffuse outlines are a common
feature of these intraformational pebbles. The claystone pebbles in (a) show very irregular outlines due to compaction
between more resistant clasts. Bars are 2 mm long.

inward (centripetal) growth style of dolomite and so low in Fe and Mn, but with narrow peaks. Three
these areas resemble cavity- or vug-fillings. However possibilities for the relative ages of type 1 and type 2
as will be seen later, their origin needs to be inter- dolomite are now evaluated.
preted with care. They are thus referred to as pseudo- (1) Type 1 dolomite replaced type 2 dolomite. This
vug-fills. seems unlikely since, during prograde diagenesis,
steadily increasing temperatures normally prevent
TYPE 1 DOLOMITE decreases in crystal size. The pattern of occurrence of
type 1 dolomite (never preferentially on pebble
Table 1 summarizes the information on this fine- margins) rules out an origin by algal or fungal
grained dolomite. It forms xenotopic mosaics which micritization of type 2 dolomite. Crystal size
occur either alone or with terrigenous impurities in diminution is possible within high strain areas (e.g.
pebbles (Fig. 3a). Type 1 mosaics also occur as twin lamellae), but such a mechanism could hardly
poorly defined and irregular patches within coarser, account for equant diffuse patches of fine-grained
type 2 dolomite (Fig. 3b, c). dolomite.
In the areas of patchy type 1 and type 2, all (2) Type 1 and type 2 formed simultaneously, the
crystals in the areas of type 2 contain abundant variation in crystal size reflecting pre-existing tex-
fluid inclusions. In the areas of type 1 dolomite, only tural variations in precursor material. The chemical
a few crystals contain abundant fluid inclusions, and differences between type 1 and type 2 could be
these crystals are never less than 15 [Lm in size. It was explained if the type 1 crystals were zoned, with rims
possible to analyse type 1 chemically only in areas in low in Fe and Mn equivalent to the dominant
which it occurred in patches in type 2. Microprobe composition of the type 2 crystals. However, the
traverses across such areas show a bimodal distri- rims of the type 1 crystals are not rich in fluid
bution of Fe and Mn (Fig. 4A and 4B). The patches inclusions and so are distinct from the type 2
of type 1 are dominantly high in Fe and Mn, areas dolomite.
low in Fe and Mn being about the same size as ( 3 ) Type 2 replaced type 1. The distinct crystals
individual crystals. The type 2 dolomite is mostly rich in fluid inclusions within the type 1 dolomite
Dolomite textures 635

Table 1. Summary description of dolomites


Type Occurrence Fluid inclusions Luminescence Fe-Mn chemistry
Type 1 Pebbles of Occasional, Uniform, very Inferred to be ferroan
5-15 pm dolostone, never arranged dull red and manganoan =
(Figs 2, 3a, silty dolostone in growth zones 0.5 wt% MnCO,,
b and c) 2 wt% FeCO,
Within pebbles, Occasional, Dull red Dominantly ferroan/
intergranular never arranged manganoan (as above).
in sandstones in growth some areas low in Fe/Mn
(as patches zones (interpreted as type 2
within dolomite crystals =
type 2 0.07 wt% MnCO,,
dolomite) 0.3 wt% FeCO,)

Type 2* Pebbles of Abundant in Abundant zones, Always chemically


15-300 pm pure and early growth dull to bright, zoned:
(Figs 2, impure stages; larger mostly red, 0.02-1 wt% MnCO,,
3b-e, 5, dolostone; crystals zoned some yellowish 0.07-3 wt% FeCO,
6, 7, 8, intergranular by variation
and 9) in sandstones in inclusion
and siltstones abundance

Type 3 Intergranular Occasional, not Dull red or 0.5 wt% MnCO,,


30-300 pm in sandstones arranged in none 0.5-2 wt% FeCO,
(Fig. 3f) growth zones
~

* Type 2 displays four growth stages (1-4); truncation textures developed at the end of stage 1 (Type B truncations)
and at the end of stage 3 (Type A truncations).

represent isolated crystals of type 2 dolomite. The TYPE 2 DOLOMITE


presence of such crystals, low in Fe and Mnj would
explain the low Fe-Mn areas in the microprobe
General features
traverses across the type 1 areas. Thus, under this
interpretation, the type 1 dolomite was originally Type 2 dolomite forms several different petrographic
ferroan and manganoan throughout. growth forms : equant polygonal mosaics, isolated
An interpretation of type 1 dolomite formation as rhombs, and intergranular material, including
a very early event is favoured by the occurrence, pseudo-vug-fills.
slightly higher in the sequence (member 3 ) , of Its characteristic fluid inclusions are mostly less
abundant unaltered fine-grained dolomite. This is than 1 pm in diameter (Fig. 3d) and each contains
identical in terms of crystal size and paucity of a vapour bubble displaying Brownian motion in a
inclusions to the type 1 dolomite as described above liquid medium. Growth zonation is often displayed
and identical in terms of the original ferroan and by variation in abundance of these inclusions. The
manganoan composition of type I dolomite inferred incorporation of inclusions appears to depend on
above. The member 3 dolomite represents a pene- crystal growth rate. Thus Valyashko (in Strakhov,
contemporaneous replacement of calcium carbonate 1970, p. 239) demonstrated that fluid-inclusion-rich
(Fairchild, 1980), so the type 1 dolomite may well zones in halite crystals grew more quickly than those
have the same origin. lacking inclusions and Spry (1969) suggested that the
In summary, types 1 and 2 dolomite are quite same rule applies to mineral inclusions in porphyro-
distinct in terms of chemistry and inclusion content blasts.
and so are unlikely to have formed at the same time. The fluid inclusions provide potential for esti-
The manner of association of type 2 with type 1 and mating the formation temperature of the type 2
the coarser crystal size of the former indicates a dolomite. They are so small that the technical
replacement of type 1 by type 2. The type 1 dolomite difficulties in determining homogenization tempera-
compares with penecontemporaneous dolomite tures are considerable, although problems of leakage
higher in the sequence. of fluids from inclusions are minimized. Rough
636 I . J . Fnirchild

Fig. 3. Photomicrographs in plane-polarized light. (a) Part of pebble of silty dolostone (mostly type 1 dolomite). Bar,
500 pm. (b) Part of dolostone pebble with patches of type 1 dolomite (dark) in coarser, type 2 dolomite. Probe analysis
shown in Fig. 4A. Bar, 500 pm. (c) Part of sandy dolostone pebble (sand grains white) containing fine-grained (type I )
intergranular dolomite partly replaced by coarser, type 2 intergranular dolomite (probe analysis shown in Fig. 4E).
Early zones of the type 2 dolomite are inclusion-rich (dark; indistinguishable from type 1 dolomite in monochrome)
whilst later zones are inclusion-free (paler) forming areas referred to in the text as pseudo-vug-fills. These areas occur in
all orientations to the bedding and are arrowed (see also Fig. 3e). Bar, 500 pm. (d) Fluid inclusions in type 2 dolomite.
Vapour bubbles faintly visible in some inclusions (arrowed). Inclusion-free dolomite zone occupies top left area. Bar,
10 pm. (e) Sandy matrix of intraformational conglomerate. Intergranular material is mostly inclusion-rich (dark) type
2 dolomite. Large, sand-free area exhibits inward growth of dolomite by luminescence and thus resembles a filled vug.
Such pseudo-vug-fills are thought to have been originally intraformational pebbles. Left-hand box shows location of
Fig. 8b, right hand box shows location of Fig. 9. Bar, 1 mm. (f) Sandstone with intergranular, type 3 dolomire. Bar,
100 pm.
Dolomite textures 637

estimates from the degree of filling (the proportion lateral coalescence of crystallites with similar optical
of the volume of each inclusion occupied by liquid orientation to form larger units (crystals).
rather than vapour, equal to less than 90% in this
case) suggest a formation temperature rather more
than 150-200C (from unpublished graphical plots
by A. H. Rankin of the data in Potter, 1977, and
Stages of type 2 dolomite growth
Potter & Brown, 1977).
A pre-tectonic age is shown by the presence of Luminescence observations and microprobe analyses
tectonic pressure shadows around dolomite crystals (Fig. 4) allow recognition of the relative time of
in claystone pebbles (Fig. 5d). This also shows that formation of the various petrographic forms of type
the dolomite crystals were more rigid than the 2 dolomite and provide evidence of the chemical
claystone. evolution of the interstitial fluids. Repeated probe
Individual quartz clast grains often penetrate analyses across a given series of zones in different
claystone clasts around their margins (Fig. 2a) crystals always show the same result: thus a corre-
because of compaction of these clasts. Dolomite lation of luminescence zones also represents a
crystals do not penetrate claystone in this way chemical correlation. The correlation of the zones is
despite their rigidity demonstrated above. This complicated by minor differences in the sequence of
favours a post-compactional age for the type 2 zones in different specimens and variations in the
dolomite. Formation before complete elimination of thickness of growth zones in different specimens or
lateral permeability is suggested by the stratigraphic in different petrographic forms such that very thin
restriction of the type 2 dolomite occurrence (Fig. 1). zones are not always resolved. Nevertheless there are
A series of rhomb-shaped luminescence zones four generally recognizable growth stages which can
characterizes most type 2 dolomite (Figs 5b, 7 and 8) be recognized by luminescence alone in the speci-
even when no zonation is visible in plane polarized mens that were not microprobed.
light (Fig. 6). Crystals were euhedral until they met, The first stage is represented by inclusion-rich
when compromise boundaries were formed. Zones dolomite which is dominantly poor in Fe and Mn,
link from one crystal to another showing that although narrow zones (5-10 pm wide) richer in Fe
growth of a particular zone was synchronous. and/or Mn do occur. A large number of luminescence
Crystals larger than about 100 pm often display zones is usually present, reflecting these chemical
undulose extinction, sometimes to the point of variations. All the finer-grained mosaics of type 2
developing sub-crystals. Although twinning is the dolomite grew entirely within stage 1 (Fig. 4A, B, D
usual response of dolomite to tectonic stress (Spry, and E). Most crystal mosaics and isolated crystals
1969), it is quite subordinate here. This implies that started to grow during this stage (Fig. 4C left, F and
the undulose extinction arose during crystal growth G). The microprobe traverse shown in Fig. 4F shows
(as can be inferred for the dolomite cements of that there were two sub-stages of growth, the first
Scherer, 1977), rather than being of tectonic origin. sub-stage exhibiting consistently low Fe; some
Crystals displaying undulose extinction also display mosaics (Fig. 4D left) grew entirely within this first
irregular crystal boundaries and growth zones show- sub-stage.
ing a complex zig-zag shape rather than the minimum Stage 2 saw the formation of inclusion-free, only
number of crystal faces. This is well displayed by the dully luminescing zones (Fig. 4C left, F and G). Fe
bright zone running E-W near the bottom of Fig. 9b and Mn are relatively low in abundance.
(crystal boundaries shown in Figs 9c and 10). There Stage 3 is represented by inclusion-rich zones,
are a number of independent units, each apparently mostly brightly luminescing (Fig. 4C left, F and G).
containing a pair of crystal faces with constant Miller Fe, and particularly Mn, are considerably higher in
indices, within each crystal. This is reminiscent of abundance than in stage 2.
the merging of crystallites of quartz and K-feldspar Stage 4 is represented by moderately or dully
overgrowing detrital grains (Waugh, 1970, 1978). luminescing zones with only one broad zone rich in
The textures also resemble the lateral coalescence of inclusions (Fig. 4C, F and G).
crystallites in stalactites described by Kendall & The pattern of growth was thus clearly complex;
Broughton (1978), particularly in terms of the most crystals nucleated during stage I , although
presence of undulose extinction and irregular crystal others did not nucleate until stage 4 (Fig. 4C right).
boundaries. These properties are dependent on the Where nucleation sites were very closely spaced,
638 1. J . Fairchild

F
A l l stoge I A l l stage I
: t e l tt PP II t P 1I I pe Patchy fine-grained
:. .i ! 1 .: (type 11 and equant
5:
5:
.t-200 p m -- - ,,
I,
c
I:

200 p m
(type 2)
Z"er,me" I8

-
All stqe 4

Zoned rhombs

w,
+I+ t 2 + - * 3 + t - 4

n (type 21 StOgeS

replacing
mudstone 3 5 0 pm lntergranuhr
G - ^ , (type 2 )
-.___
, -.-~
>,.,"I___ <-*
I
- _ I
.
-..
spec men za

All Itoge I
I
Equant
replacive
. .-
(type 21 H: --- : 3.. >,.
I Intergranular
.. ... 2
-., -.; (type 31

Sper~man 27

A l l stage I

Intergranular (type 2)
t"iE
::2s
ipe: llSI 11
.:j
a 01

Fig. 4. Microprobe analyses (scans) of Fe and Mn in dolomite (A-H) and relation to inclusion and luminescence zones
Mn
Fe

t--,q
,
l i r P I

f yen;y inllUI,anr

(C, F, G); note key at lower right of figure. The horizontal scales on C , F and G are inexact since these diagrams were
constructed from scans across several crystals; the scale indicates the minimum distance (equivalent to a section at a
high angle to the zones). Fe and Mn had to be scanned on separate lines so their distribution could only be superimposed
in the case of the larger crystals (C, F and G ) . Spot checks show that Fe and Mn vary sympathetically in A, B and D,
and to some extent in E. In the case of H, the diagram shows the results of analysis of five crystals, two of which are
zoned with the rims richer in Fe; each crystal is 50-150 pm wide. Specimens 2, 27 and 28 are intraformational con-
glomerates and specimen 29 is a sandstone. Relevant photomicrographs: A, Fig. 3b; C, left, Fig. 5a; D, left, Fig. 6a
and b ; D, right, Fig. 6c and d ; E, Fig. 3c; F, Fig. 7 ; G, Figs 3e, 8 and 9.

growth ceased in stage 1. Conversely, a more wide REPLACIVE CRYSTALS I N


or less regular distribution of nucleation sites allowed C L A Y S T O N E S A N D REPLACIVE
growth to continue t o stage 4. Different petrographic DOLOMITE MOSAICS
forms grew a t different rates (cf. Fig. 4C left and G).
Also the relative rates of growth of crystals a t Type 2 dolomite forms euhedral crystals and poly-
different times were variable. Thus for example gonal mosaics within claystone pebbles (Figs 5 and
growth stage 3 was relatively thicker compared with 6a). As only pebbles of claystone contain dolomite
growth stage 2 in some samples (e.g. Fig. 4F) than crystals, these crystals clearly formed by replacement
in others (e.g. Fig. 4G). These features will be of claystone. A complete range from unreplaced to
discussed further later. nearly completely replaced pebbles may occur, even
Dolomite textures 639

Fig. 5. Partial replacements of claystone by type 2 dolomite. (a, b) Zoned dolomite rhombs within claystone pebble
under transmitted light and by luminescence respectively. Probe scan lines (analysis shown in Fig. 4C, left) running
NE-SW are visible. Crystals have a dark line of insolubles, beyond which is non-luminescing dolomite forming a tectonic
pressure shadow. Bars, 100 pm. (c) Claystone pebble partly replaced around its margins by clear (type 2 or 3) dolomite
(plane polarized light). Bar, 500 pm. (d) Zoned type 2 dolomite which has replaced claystone (plane-polarized light).
Dark line of insolubles visible at the margins of diagenetic (cf. tectonic) growth. Tectonic pressure shadows (sense of
Spry, 1969) of dolomite are visible at top and bottom of each crystal with characteristic intergrowth with mica (best
displayed at base of photograph) accompanied by dissolution of dolomite on N-S-trending margins in the photograph.
Occasional aligned trains of mica are the only other evidence of tectonic deformation in these beds. Bar, 100 pm.

within a single thin section (e.g. fig. 4 of Fairchild, more slowly than intergranular rhombs in the same
1977). Fig. 5 illustrates partial replacements of rock (cf. Fig. 4C left and F) since all the growth
claystone pebbles by relatively large dolomite stages are thinner in the former.
crystals, and Fig. 6a shows near complete replace- Unconnected with claystones are pebbles of 30-
ment of a claystone pebble by smaller, polygonal 100 pm mosaics of type 2 dolomite which lack dark
crystals of dolomite. In each case there is dark insoluble matter at their crystal boundaries (Fig. 6c).
insoluble matter (oxides and some mica) at the All transitions exist from such dolomite to similar
margins of crystals (except where modified by later mosaics containing patches of type 1 dolomite (Fig.
tectonic solution transfer, Fig. 5d), indicating a 3b), to mosaics of type 1 dolomite, with (Fig. 3a) or
displacive component as well as the main replacive without floating quartz silt. The type 2 dolomite here
action. Despite the fact that not all of the consti- clearly must be replacive. The evidence is consistent
tuents of the claystone could be dissolved as growth with replacement of type 1 dolomite in each case,
proceeded, dolomite crystals in claystone always since this is the only other material present, although
developed crystal faces until they met others of their complete replacement of another carbonate or even
own kind. The dolomite rhombs in claystone grew non-carbonate material cannot be excluded,
640 I . J . Fairchild

Fig. 6. Equant, polygonal type 2 dolomite in pebbles. (a, b) Dolomite that has replaced claystone seen under transmitted
light and by luminescence respectively. Concentration of insolubles along crystal boundaries is visible in (a). Internal
rhombic zones are visible in (b). Probe analysis shown in Fig. 4D, left. Bars, 200 pm. (c, d) Dolomite interpreted as a
replacement of type 1 dolomite seen under transmitted light and by luminescence respectively. Extensive linking of
later zones is visible in (d) with one area (labelled X) showing only late zones. Probe analysis shown in Fig. 4D,right.
Bars, 100 pm.

adjacent to them (Figs 7 and 8). (4)The margins of


Intergranular dolomite
sand grains are fretted by dolomite (Figs 7 and 9).
The intergranular type 2 dolomite in sandstones Even the late zones of the dolomite crystals show
and in sandy dolostone pebbles varies in crystal size this feature.
from 15 to 300 pm, although less variable in crystal These observations demonstrate a replacive origin
size in small areas. The following points are particu- for a t least the earlier zones, but would not rule out
larly crucial for understanding its genesis. (1) The the formation of cavities during the later stages of
sand grains are always completely surrounded by dolomite growth if partial etching of quartz clasts
dolomite which typically occupies about 70% of a also occurred in contact with pore fluid. The evidence
given area. (Figs 2a, 2b, 3c, 3e and 7). (2) Ill-defined is consistent with all the intergranular material prior
patches of type 1 dolomite are often present within to type 2 dolomite formation being type 1 dolomite,
the type 2 dolomite (e.g. Fig. 9c top). (3) I n most perhaps in the form of compressed primary sand
cases, including all examples where type 1 dolomite sized grains. However, the former presence of, for
is absent, the points of origin of the dolomite example, another carbonate again cannot be ruled
crystals are always between sand grains rather than out.
Dolomite textures 641

Fig. 7. Intergranular dolomite in sandstone. Inclusion and luminescence zones, and microprobe analyses are shown in
Fig. 4F. (a, b) Transmitted light and luminescence respectively. Arrowed crystal shows probe scan lines (WNW-ESE)
and type A truncation surface cutting the composite bright zone. c = quartz and microcline clasts (the latter luminesce
uniformly in b). Bars, 200 pm.

Pseudo-vug-fills of dissolution of the pebble and cavity-filling by late


The origin of dolomite in the pseudo-vug-fills pro- dolomite zones, or for slow ordered growth by
vides distinct problems. Pseudo-vug-fills are nearly dolomite giving rise to a complete replacement
always elongate and occur at all angles to the lacking inclusions. This is discussed further in the
bedding. They are often an order of magnitude larger next section.
than the sand-grain-supported interstices (Fig. 3e).
There is often some type 1 dolomite present within
their margins (e.g. Fig. 9c, top), but it is only Truncation textures and pseudo-vug-fills: cementing
associated with the inclusion-rich early growth stage or replacement origin?
of the type 2 dolomite. Fig. 8 shows that early The type 2 intergranular dolomite displays two kinds
zones of crystals (forming the cores of rhombs) are
of intracrystalline truncation surfaces, here named
only present near the margins of pseudo-vug-fills, types A and B, separating inclusion-rich from
and conversely that the latest zones occur only near inclusion-free dolomite. Type A are described first
centres. The pseudo-vug-fills are thus the products
as they are the simpler geometrically.
of inward (centripetal) growth of dolomite.
An origin of pseudo-vugs as primary cavities or as
vugs formed by gas generation or differential lithi- Type A truncation surfaces
fication seems most unlikely. This is because of their In Fig. 7b, the dully luminescing initial zone of
large size, elongate shape at all angles of bedding and growth stage 4 (Fig. 4F) clearly truncates earlier
occurrence solely within arenites and rudites. The zones along a very irregular line, nearly cutting out
absence of nodular shapes or crystal pseudomorphs growth state 3 altogether in places in the arrowed
rules out any origin connected with evaporites. The area of Fig. 7b. Identical features are shown in other
presence of some fine-grained dolomite and the specimens (Fig. 8a), where they are displayed all the
shape of the pseudo-vug-fills rules out an origin way around the pseudo-vug-fills. Fig. 8b shows a
by formation of fracture cavities. However, pebbles very well developed example of truncation where, in
occur in the same rocks with similar sizes, shapes and the bottom left area, all previous growth stages are
orientations to the pseudo-vug-fills. It is thus truncated by the first zone of growth stage 4. Such
probable that the latter were originally tabular truncations are seen in a number of other specimens,
pebbles containing some type 1 dolomite, at least but the degree of truncation is not usually so pro-
near their margins. The interior portions of pseudo- nounced. The truncation surfaces never coincide with
vug-fills, lacking inclusions of fluid, or of type 1 crystal boundaries.
dolomite could be regarded as providing evidence Two hypotheses are now put forward for the
642 I . J. Fairchild

Fig. 8. Photomicrographs of luminescence of parts of the pseudo-vug of Fig. 3e showing zonal sequence and type A
truncations. Quartz clasts are non-luminescing. (a) Truncation surfaces are visible, e.g. in the arrowed area. Probe scan
lines (analysis of Fig. 4G) run NE-SW in the centre of the photograph. Bar, 100 pm. (b) Location of this photograph
with respect to the pseudo-vug is shown in Fig. 3e. The brightly luminescing zones are cut out (centre) by a later zone
and are absent throughout the bottom left region above and to the left of the large quartz clast. This is a type A trun-
cation texture. Early dolomite zones (centres of rhombs) are restricted to the margins of the pseudo-vug-fill (cf. Fig. 3e).
Growth was thus inward (centripetal). Bar, 200 pm.

formation of these textures. The dissolution-over- in cements described by Meyers (1978) which he
growth hypothesis proposes that the crystals were attributed to dissolution between phases of cement
partly dissolved adjacent to a cavity and later over- precipitation. Fracturing was sometimes associated
grown in optical continuity (Fig. 11D, E and K). with this dissolution. However the textures now
Although this is the most obvious and familiar being described seem much less irregular than these
possibility, an alternative should be considered examples except when solution was extensive(Fig. 5b.
bearing in mind that the earlier growth stages of the of Meyers, 1978). The type A truncations are
intergranular crystals have been shown to be extremely irregular with very sharp re-entrants, even
replacive in origin. Thus the solution film hypothesis where the amount of truncation is very small (Fig.
proposes that the crystals were dissolved and 7b) and are never associated with fracturing. Disso-
reprecipitated by a solution film migrating towards lution surfaces might be expected to be more smooth
the older zones, in doing so changing their chemical (cf. e.g. Graf & Lamar, 1950). Scherer (1977) con-
composition to that of the stage 4 zone immediately sidered that an example of irregular dissolution of
following the truncation surface (Fig. 11H, I/J and dolomite that he discovered was a special phenome-
K). non caused by high Eh conditions under which Fe
Considering the first hypothesis, the textures seem in the lattice was oxidized creating local weaknesses
broadly geometrically comparable with truncations and hence non-uniform solution ;there is no evidence
Dolomite textures 643

from the example now being described that such Fig. 3e (Figs 9 and 10) and elsewhere in the same
conditions existed. Another point is that the onset specimen.
of dissolution rather than precipitation of dolomite In plane-polarized light a pattern of anastomosing
requires a drastic change in solution chemistry, yet areas of inclusion-rich dolomite separating domi-
there is no circumstantial evidence (such as the nantly inclusion-free areas (Figs 9a and 10) can be
association with unconformities in the case of seen. Luminescence reveals no distinct zones within
Meyers, 1978) to suggest why this should happen. the inclusion-rich dolomite (Fig. 9b) presumed to be
The second hypothesis, involving a moving of growth stage 1. The lack of obvious growth zones
solution film (e.g. as suggested by Kinsman, 1969) means that the presence of truncations ha3 to be
would involve a similar process to that operating deduced rather than simply observed. The fact that
when aragonite is replaced by calcite without loss of growth of the inclusion-free dolomite was always
crystal texture (Bathurst, 1971, p. 347) and possibly away from adjacent inclusion-rich dolomite is shown
that operating when Mg is lost from Mg-calcite by the shape of the crystallite faces. Sharp re-entrants
(Bathurst, 1971, p. 337). The replacement would always face towards the inclusion-rich areas, crystal-
occur at the onset of, or during the precipitation of, lite faces always face away from the inclusion-rich
the zone following the truncation surface (Fig. I11 areas. This growth took place in two distinct ways.
and J). The pattern is always of inclusion-rich Firstly there is the usual growth into the pseudo-vug
dolomite being truncated by inclusion-free dolo- (with a prominent narrow (red) bright zone of
mite; a lower solubility of the latter could provide growth stage 3). This is seen near the bottom of Fig.
the driving force for the replacement. Uncertainty as 9b (line of downward-pointing arrows in Fig. 10).
to the effectiveness of this driving mechanism is a Secondly, in the centre of Fig. 9b there are a number
clear weakness in this hypothesis. It is difficult to of small areas of centripetal growth containing a
assess whether the form of the truncation surface is thicker bright zone (seen in colour to be a triplet
consistent with a solution film because preserved red-yellow-red) also representing growth stage 3 (see
solution films are uncommon. Pingitore (1976) did later for evidence for this). These small areas are
record such features, but did not indicate the degree called small centripetal fills to distinguish them from
of irregularity of the solution film surface. He the pseudo-vug-fills. The remarkable thing about the
indicated that a narrow, texture-retentive film, such small centripetal fills is that in general a single
as that required here could operate below the vadose crystal spans their width (Fig. 10). Normally, in
zone as long as diagenesis was not very rapid. other rocks when growth zones indicate convergent
Wardlaw, Oldershaw & Stout (1978) illustrated growth, a crystal boundary will lie at the centre of
extremely irregular calcite-aragonite contacts com- the area which has been grown into, representing the
parable with the type A truncations, but these line where the converging crystals met. This is not
authors took the view that it was conceivable that the case here. Growth zones within single crystals are
solid-state inversion rather than a solution film normally divergent, here they are convergent !
process had operated in their case. How can this be explained?
One certain way of disproving the dissolution- One possibility is that initially a number of small
overgrowth hypothesis would be to find truncations crystals of various optical orientations grew in-
completely surrounding growth stage 3 of dolomite wards, forming the luminescence zones, and estab-
crystals isolated within claystone, since such crystals lishing the existence of the small centripetal fills.
could never have been suspended in mid-air. Later they would have been replaced by the large
However, as the zones in such crystals are relatively crystals now seen to be spanning the small centri-
thin (Fig. 5a) it is not possible to distinguish trun- petal fills. However the Miller indices of crystal
cation of the latest of three bright zones of growth faces seen by luminescence appear to be constant
stage 3 (as Figs 4G and Sa) from loss of resolution of within each of the large crystals as was pointed out
this zone because it is so thin! The two hypotheses earlier. There are thus no grounds for supposing
thus both remain applicable. that variably orientated crystals were replaced. In
any case, if this later process were one of wet
boundary (solution film) migration then the earlier
Type B truncation textures zonal sequence could not have been preserved.
These form a number of smooth, closed surfaces Neither does a dry process of grain growth seem
distributed along the margins of the pseudo-vug of possible, since the crystals contain distinct areas of
644 I . J . Fairchild
Dolomite textures 645

Before proceeding to a discussion of the relative


merits of the hypotheses, two minor points need t o
be cleared up. The first point relates to the lowest
(dashed and dotted) line drawn in o n Fig. 9a. This
represents the boundary between inclusion-rich and
inclusion-free dolomite (i.e. between growth stages
1 and 2) around the margins of a pseudo-vug, rather
than around a small centripetal fill. This stage 1/
stage 2 interface is demonstrably one of truncation
within the small centripetal fills, but truncation
cannot be proved for thedashedand dotted line. This
is because individual crystals d o not span the width
of pseudo-vug-fills (unlike the small centripetal fills)
Fig. 10. Explanatory line diagram of the area shown in so the evidence necessary t o establish the presence
Fig. 9. Symbols as for Fig. 1 1 . The lower part of the area,
with downward-pointing arrows represents crystal of truncation is not present.
growth into the pseudo-vug. The central area shows a The second minor point relates to the assumption
number of areas (bounded by the type B truncations made above that the bright zones in the pseudo-vug-
indicated in Fig. 9a) of inward growth: these are the fills and the small centripetal fills were synchronous
small centripetal fills. For clarity two features are omitted :
(growth stage 3). They appear somewhat dissimilar
crystal boundaries within dolomite near the top of the
field of view (because of the presence of fine-grained at first sight since the zones in the small centripetal
dolomite there) and inclusions within dolomite showing fills are thicker. However this is simply explained by
bright marker zones (growth stage 3 ; the relation of dolomite in the small centripetal fills having grown
inclusion zones to luminescence zones is however shown more quickly than in the pseudo-vug-fills. Each zone
in Fig. 4G). These omissions are also made in Fig.
1 IA-C, F and G. is thus wider in the small centripetal fills, several very
thin zones (including the central yellow part of the
fluid inclusions. There is a n alternative possibility main bright zone) being visible here, but not
which can explain the textures of Fig. 9: this requires resolved in the thinner sequence in the pseudo-vug-
truncation surfaces t o be present. There would first fills. T h e sequence o f inclusion-rich and inclusion-
have been a stage of growth of inclusion-rich free zones (growth stages 2-4 of Fig. 4G) is the same
crystals spanning the areas now occupied by the in both.
small centripetal fills. If each of these crystals then Compared with the type A textures, the type B
suffered internal dissolution, they could be recon- truncation surfaces are rather smooth (Figs 9a and
structed by centripetal syntaxial growth. That is, the lo), consistent with being the walls of dissolution
original crystals acted as a template t o determine cavities (e.g. Graf & Lamar, 1950), but there is n o
the optical orientation of the crystals making u p the reason why a solution film could not have had such a
small centripetal fills. The outer margins of the small smooth form originally.
centripetal fills are represented by a boundary Both hypotheses suffer from difficulties in moving
between inclusion-rich dolomite outside and domi- fluid and/or ions t o the sites of re-growth since these
nantly inclusion-free dolomite inside (indicated by sites would be isolated. The problem is more acute
dashed lines in Fig. 9a). The original inclusion-rich than the usual one in carbonate petrology in plugging
crystals were truncated along this boundary : these the last remaining pores in limestones since the small
boundaries are the type B truncation surfaces. centripetal fills occur entirely within single crystals :
Having established the existence of truncations, material diffusing along crystal boundaries alone
both the dissolution-overgrowth hypothesis and the would not have reached the sites of growth. How-
solution film hypothesis can be invoked to explain ever, as the crystals seem t o have formed from co-
them. This is demonstrated in the caption to Fig. 11. alescing crystallites and exhibit undulose extinction,
Fig. 9. (a) Transmitted light; (b) luminescence; (c) crossed polars. Part of the margin of the pseudo-vug-fill of Fig, 3e
showing type B truncations. The trace of the truncation surfaces are shown dashed in (a). They enclose dominantly
inclusion-free areas which are the small centripetal fills. The dashed and dotted line is the boundary between growth
stages 1 (inclusion-rich) and 2 (inclusion-free) around part of the pseudo-vug-fill; it is not possible to tell whether
growth stage 1 has been truncated here or not. Patchy fine-grained (type 1) dolomite is visible (e.g. where arrowed) in
the upper part of (c). Further explanation is given in Figs 10 and 11 and the text. Bars, 200 pm.
646 I . J . Fairchild

.Y Crystal boundaries

Inclusion-rich dolomite
C Quartz or microline clast
V vu9
Tabular pebble being
X
replaced by dolomite

4 (( Growth directions of dolomite


I , and marker bright zone (s)

Fig. 11. Interpretations of type B and type A truncations. (A) Growth of inclusion-rich type 2 dolomite into a tabular
pebble (pseudo-vug) from around the pebble margins. (B-E) Dissolution-overgrowth hypothesis. (B) Vugs are formed
by leaching within pre-existing dolomite crystals. The tabular pebble may or may not be leached out at this time. The
leached surfaces are the type R truncations. ( C ) Residual dolomite is overgrown in optical continuity to form small
centripetal fills as well as to further fill the pseudo-vug (bottom). (D, E) Following growth of a second bright zone
within the dolomite growing into the pseudo-vug, leaching causes corrosion of the crystal surface of the inclusion-rich
dolomite (type A truncation surface). Later, growth in optical continuity takes place filling the remaining cavity (K).
(F-J) Solution film hypothesis. (F, G) A series of inwardly-moving solution films dissolve inclusion-rich dolomite and
precipitate inclusion-free dolomite within the small centripetal fills. At the same time, growth continues into the pseudo-
vug (bottom). The initial configurations of the solution films are the type B truncation surfaces. (H, 1, J) Following
growth of a second bright zone within the dolomite growing into the pseudo-vug (H), a solution film (moving upwards
in the diagram) dissolves inclusion-rich dolomite and precipitates inclusion-free dolomite. The type A truncation
surfaces are the final positions of the solution films. Migration of the solution film could take place before any further
downward growth of dolomite into the pseudo-vug (I) or simultaneously with such downward growth (J). Later growth
in optical continuity takes place completing the pseudo-vug-fill (K).

surfaces of high lattice misfit could have pro- tures (in excess of 150-200C) suggested earlier t o
vided diffusion channelways, especially if crystallite have been involved.
coalescence had not been completed. This would A n argument against one version (V a t base of
have been assisted by the relatively high tempera- Fig. 11B and C ) of the dissolution-overgrowth
Dolomite textures 647

hypothesis is the fact that the growth rate was faster immediately below member 1, or laterally adjacent
within the small centripetal fills (isolated within to it where it wedges out to the west suggests a source
crystals) than into the main pseudo-vug (large of dolomitizing ions from the centre of the deposi-
cavity). This does not apply to the other version (X tional basin to the east. The sharp top to the
of Fig. 11B and C) which does not postulate disso- dolomitic sediments could have arisen because of a
lution of the pebble occupying the pseudo-vug at the low permeability of the overlying silica-cemented
time the type B truncations formed. sandstones (member 2). The distribution of silica
In summary, there is insufficient evidence to rule cementation would ultimately have been determined
out either hypothesis for the formation of the type B by the presence of clasts of type 1 dolomite in
truncation textures. member 1. These became squashed between quartz
clasts on compaction, thus inhibiting silica cemen-
tation. The nucleation of dolomite crystals at the top
TYPE 3 DOLOMITE of member 1 was probably aided by the fact that
there was so much fine-grained dolomite already
Type 3 dolomite (Fig. 3f; 30-300 pm) is inclusion- present as clasts at this horizon.
free and occurs as intergranular material in sand- Control of the composition of type 2 dolomite
stones. It is usually anhedral, but crystals sometimes crystals by factors external to the sediments examined
show some rhombic faces. Unlike type 2 dolomite is indicated by the fact the same growth stages can
it typically forms an inconzplete intergranular border be recognized in dolomite replacing claystone as for
to clasts. It corrodes quartz clasts, but only to the the other examples of type 2 dolomite. The chemical
extent of an irregular fretting of grains. It shows an zones of the type 2 dolomite show extremely complex
antipathetic relationship to sand-sized clay clasts. zonation (Fig. 4), Fe and Mn showing at different
Occasionally, it occurs as the only intergranular times, sympathetic, antipathetic and uncorrelated
material in millimetre-sized patches independent of co-variation. Such intricate crystal zonation seems
the sedimentary lamination and surrounded by sand characteristic of carbonate cementation (Evamy,
containing clay clasts as the only intergranular 1969; Meyers, 1974) and replacement (Katz, 1971;
material. this work) in the sub-surface. This presumably
Microprobe analyses of type 3 dolomite in one relates to fluctuating availability of ions at source,
sample (Fig. 4H) showed it to be ferroan and and in the case of Fe and Mn, fluctuating Eh en
manganoan, and staining of other samples confirms route to the depositional site (Meyers, 1974, 1978),
that type 3 dolomite is always ferroan. Two growth although the ultimate cause of these variations is not
zones are present in some crystals. known.
The occurrence in discrete patches and its anti- The same four main stages of growth can be
pathetic relationship to sand-sized clay clasts is recognized in all the samples examined by lumines-
evidence for an origin, at least in part, by replace- cence petrography (all from Section B, except one
ment of clay clasts. The similarity in terms of sampled 3 km inland to the west). Minor differences
chemistry and lack of inclusions between type 3 in zonation do occur within these stages, but these
dolomite and growth stage 4 of type 2 dolomite do not clearly relate to specific horizons. The most
suggests that the two may have been synchronous. clear difference between samples is in the relative
thicknesses of different growth stages (e.g. cf.
differences in relative thicknesses of growth stages
DISCUSSION
2 and 3 between Fig. 4 F and G). This indicates
temporal variations in the relative growth rate of
Origin and chemistry of types 2 and 3 dolomite
crystals in different specimens. Minor differences of
The stratigraphic restriction of these dolomite types this sort occur even within different areas of the same
coupled with the post-compactional age (at tem- hand specimen. This feature may possibly be a
peratures higher than 150-200C) indicate that they characteristic of replacive crystals contrasting with
do not fit any of the normal dolomitization models the constant relative growth rates expected from
which refer to near-surface conditions or association cements growing in interconnected pores.
with fractures. The types 2 and 3 dolomite seem to The later stages of dolomitization (type 3 dolomite
represent an example of dolomitization by deep and growth stage 4 of type 2 dolomite) are charac-
burial diagenesis. The lack of carbonate above or terized by crystals which are inclusion-free. This
648 I . J . Fairchild

presumably reflects slower growth under conditions 1971) arise because crystals tend not to develop
of reduced supersaturation compared with earlier crystal faces and do not all grow at the same time.
growth stages. The later stages are also characterized Large numbers of crystallites are a feature of
by crystals with relatively high Fe and Mn content, dolomite forming after the type B truncations had
similar to the type 1 dolomite. In the growth of developed (Fig. 9b). Thus the presence of coalescing
garnet porphyroblasts (Edmunds & Atherton, 1971) crystallites, rather than being a normal feature of
the incorporation of minor elements may sometimes dolomite growth, could be evidence for intermittent
be controlled by their diffusion rates, minor element growth, a new phase of growth being marked by the
content increasing when growth slows. In type 2 development of numerous crystallites. Again this
dolomite, however, decreasing quantities of in- could apply to cements or replacements.
clusions never coincides with increasing Fe and Mn The degree to which replacement textures can
(Fig. 4C, F and G) which suggests that diffusion- mimic those of cements would depend on the
limited supply was not an important factor. However locations of nucleation sites during replacement. In
there is an intrinsic explanation for the high Fe and Fig. 6b the nucleation sites were very regularly
Mn content of later growth stages. If it is accepted distributed and closely spaced so that the resulting
that type 2 dolomite replaced type 1 dolomite then texture could not possibly be confused with a cement.
during this process much Fe and Mn would have Fig. 6d shows an area (labelled X) which contains
entered the solution, and unless the system was only late zones and thus resembles a cavity-filling.
completely open some of this would have been No such features are present in fine-grained dolo-
incorporated in the type 2 dolomite. Slower flow mite pebbles unreplaced by type 2 dolomite. Area X
rates, that is less introduction of allochthonous is thus unlikely to represent a cavity-filling but rather
material, causing a more closed system, would favour seems to be an example of local inhibition of
a composition of type 2 dolomite close to that of nucleation of crystals. Also, the pseudo-vug-fills
type 1 dolomite. This is the case for the later stages were proposed to be replaced pebbles and yet the
of dolomitization. manner of growth was always inwards as for a
cement. If the replacement hypothesis is true, this
requires development of nucleation sites at the
Similarities between cementation and replacement by
pebble margins and inhibition of nucleation within
dolomite
the pebble. Fig. 5c shows a claystone pebble
The foregoing discussion raised problems in deciding encroached upon around its margins by dolomite
whether or not dolomite had formed in pore space. which shows that such restricted nucleation is
These features are highlighted below. possible. This demonstrates that considerable cau-
In terms of crystal growth patterns there may be tion must be applied when interpreting dolomite
considerable similarities between cementing textures textures as cements.
and those of euhedral replacive crystals. The type 2
dolomite was apparently able to form crystal faces as
easily as if growing into a cavity. Katz (1971) con- CONCLUSIONS
cluded that the material being: replaced must be soft
for this to happen, but this is certainly not necessary Two distinct periods of dolomitization are recog-
as the crystals need not force aside the matrix, but nized in member 1: an early, ? penecontemporaneous
instead it may be dissolved at their periphery as one (type 1 dolomite), and a later one (types 2 and
growth proceeds: that is, there is a chemical, not a 3 dolomite), a product of burial diagenesis. This later
physical, control. period displays a number of remarkable and novel
The crystal boundaries formed should be identical features. Replacement of claystone occurred, crystal
to those of cements since, in the examples of replace- faces always being present. Crystal growth rates in
ment described here, adjacent crystals show the same claystone were slower than in crystals replacing
growth zones and so grew synchronously. The carbonate in the same rock. Inclusion-rich zones and
crystal boundaries are in places irregular because of chemical zones of crystals can be correlated regard-
the development of coalescing crystallites, but this less of the nature of the material being replaced.
can be a feature of cavity-fills also (Kendall & Crystal growth was interrupted at two stages and
Broughton, 1978). The normal neomorphic textures crystals dissolved or cannibalized. On the first
of calcite with irregular crystal boundaries (Bathurst, occasion the inside of crystals was affected, giving
Dolomite textures 649

rise t o previously undescribed textures (type B EVAMY, B.D. (1969) The precipitational environment and
truncations). O n the second occasion the free growth correlation of some calcite cements deduced from
artificial staining. J . sedim. Petrol. 37, 1204-1215.
surfaces of crystals were affected (type A trun- FAIRCHILD, I.J. (1977) Phengite spherules from the Dal-
cations). T h e rocks became closed t o F e and M n in radian Bonahaven Formation, Islay, Scotland :
the final stages of dolomite growth. glauconitized microfossils? Geol. Mag. 114, 355-364.
Because of the euhedral form of dolomite crystals, FAIRCHILD, I.J. (1978) Sedimentation and post-deposi-
tional history of the Dalradian Bonahaven Formation
the distinction of cementing and replacement tex- of Zslay. Unpublished Ph.D. thesis, University of
tures is much more difficult than for calcite. Replace- Nottingham.
ment can follow the same ruler of competitive FAIRCHILD, I.J. (1980) Sedimentation and origin of a late
growth a s cementation. Restriction of nucleation of Precambrian 'dolomite' from Scotland. J . sedini.
crystals t o the margin of a n area t o be replaced Petrol. SO, 423-446.
GRAF,D.L. & LAMAR, J.E. (1950) Petrology of Fredonia
could lead t o a complete simulation of a vug-filling Oolite in southern Illinois. Bull. Am. Ass. Petrol.
texture by replacive crystals. Truncation textures Geol. 34, 23 18-2336.
within crystals d o not necessarily indicate dissolution HARRIS,A.L. & PITCHER,W.S. (19751 The Dalradian
adjacent t o pore space. Conversely the presence of Supergroup. In : A Correlation of Precambrian Rocks
o f the British Isles. Spec. Rep. geol. Soc. Lond. 6,
irregular crystal boundaries does not demonstrate 52-75.
replacement. Perhaps the most useful line of evi- KATZ,A. (1971) Zoned dolomite crystals. J. Geol. 79,
dence relates t o the degree of correlatability of 38-51.
crystal zones, a more widespread correlation being KENDALL,A.C. & BROUGHTON, P.L. (1978) Origin of
expected from cements as associated rock per- fabrics in speleothems composed of columnar calcite
crystals. J . sedim. Petrol. 48, 519-538.
meabilities should be higher than when replacement KINSMAN, D.J.J. (1969) Interpretation of Sr?+ concen-
is occurring. Comparing any two samples in this trations in carbonate minerals and rocks. J. sedim.
study, the relative rate of growth of replacive Petrol. 39, 486-508.
dolomite crystals in the two samples is not constant: MARSHALL, D.J. (1978) Suggested standards for the
reporting of cathodoluminescence results. 1. sedim.
this alone may prove to be a diagnostic criterion of Petrol. 48, 651-653.
replacement. MEYERS, W.J. (1974) Carbonate cement stratigraphy of
the Lake Valley Formation (Mississippian), Sacra-
mento Mountains, New Mexico. J . sedim. Petrol. 44,
ACKNOWLEDGMENTS 837-861.
MEYERS, W.J. (1978) Carbonate cements : their regional
distribution and interpretation in Mississippian
The author gratefully acknowledges receipt of a limestones of southwestern New Mexico. Sedi-
N.E.R.C. studentship whilst a t Nottingham Uni- mentology, 25, 371-399.
versity. D r P. K. Harvey and P. O'Byrne gave MOBERLY, R. (1968) Composition of magnesian calcites
of algae and pelecypods by electron microprobe
instruction and advice o n microprobe operation. analysis. Sedimentology, 11, 61-82.
D r J . A. D. Dickson provided stimulating super- MOBERLY, R. (1970) Microprobe study of diagenesis in
vision of this study and did much of the work calcareous algae. Sedimentology, 14, 113-1 23.
involved in developing the microprobe technique. PIERSON, B.J. (1977) The control of'cathodo-luminescence
Technical facilities were provided by the Universi- in dolomite by iron and manganese. Unpublished M.S.
thesis, University of Kentucky.
ties of Nottingham and Cambridge. The manu-
PINGITORE, N.E. (1976) Vadose and phreatic diagenesis:
script was much improved a t several stages by the processes, products and their recognition in corals.
criticisms of Drs C. V. Jeans, J. A. D. Dickson, P. F. J. sedim. Petrol. 46, 985-1006.
Friend, R. G. C. Bathurst, two anonymous referees, POTTER,R.W. (1977) Pressure corrections for fluid
and Dr J. D. Hudson. inclusion homogenization temperatures based on the
volumetric properties of the system NaCI-H,O. US.
geol. Surv. J . Res. 5, 603-607.
POTTER,R.W. & BROWN,D.L. (1977) The volumetric
REFERENCES properties of aqueous sodium chloride solutions from
0" to 500C at pressures up to 2000 bars based on a
BATHURST, R.G.C. (1971) Carbonate sediments and their regression of available data in the literature. Bull.
diagenesis. In : Developments in Sedimentology, 12. US.geol. Surv. 1421C, 36pp.
Elsevier Publishing Co., Amsterdam. REED,S.J.B. (1973) Principles of X-ray generation and
EDMUNDS, W.M. & ATHERTON, M.P. (1971) Poly- quantitative analysis with the electron microprobe.
metamorphic evolution of garnet in the Fanad In: Microprobe analysis (Ed. by C . A. Anderson).
aureole, Donegal, Eire. Lithos, 4, 147-161. Wiley & Sons, New York.
650 I . J. Fairchild

SCHERER, M. (1977) Preservation, alteration and multiple SPRY,A. (1969) Metamorphic Textures. Pergamon Press,
cementation of aragonitic skeletons from the Cassian Oxford.
Beds (U. Triassic, Southern Alps) : Petrographic and STRAKHOV, N.M. (1970) Principles of Lithogenesis, vol. 3
geochemical evidence. N . Jb. Geol. Palaont. Abh. 154, (English translation). Oliver & Boyd, Edinburgh.
213-262. WARDLAW, N., OLDERSHAW, A. & STOUT,M. (1978)
SIPPEL,R.F. & GLOVER,E.D. (1965) Structures in car- Transformation of aragonite to calcite in a marine
bonate rocks made visible by luminescence petro- gasteropod. Can. J. Earth Sci. 15, 1861-1866.
graphy. Science, 150, 1283-1287. WAUGH,B. (1970) Formation of quartz overgrowths in
SOMMER, S.E. (1972) Cathodoluminescence of carbon- the Penrith Sandstone (Lower Permian) of northwest
ates. I. Characterization of cathodoluminescence from England as revealed by scanning electron microscopy.
carbonate solid solutions. Chem. Geol. 9, 257-284. Sedimentology, 14, 309-320.
SPENCER, A.M. & SPENCER, M.O. (1972) The late Pre- WAUGH,B. (1978) Authigenic K-feldspar in British
Cambrian/Lower Cambrian Bonahaven Dolomite of Permo-Triassic sandstones. J. geol. SOC.Lond. 135,
Islay and its stromatolites. Scott. J. Geol. 8, 269-282. 51-56,

(Manuscript received 8 April 1979; revision received 19 November 1979)

Potrebbero piacerti anche