RECENT SEDIMENTATION
IN NORTHERN
CARDIGAN BAY, WALES

J. ROBERT MOORE

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
MINERALOGY Vol. 2 No. 2
LONDON: 1968

RECENT SEDIMENTATION
IN NORTHERN CARDIGAN BAY, WALES

BY

J. ROBERT MOORE

University of Wisconsin

Pp. 19-131 ; Plates 2-6; 47 Text-figures

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
MINERALOGY Vol. 2 No. 2
LONDON : 1968

THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), imstituted in 1949, 1s
issued in five series corresponding to the Departments
of the Museum, and an Historical serves.

Parts will appear at irregular intervals as they become
veady. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 2, No. 2 of the Mineralogy
series. The abbreviated titles of periodicals cited
follow those of the World List of Scientific Periodicals.

World List abbreviation
Bull. Br. Mus. nat. Hist. (Miner.).

© Trustees of the British Museum (Natural History) 1968

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 2 February, 1968 Price £3

RECENT SEDIMENTATION
IN NORTHERN CARDIGAN BAY, WALES

By J. ROBERT MOORE

CONTENTS
Page
I. INTRODUCTION - - - : : : - : : 22
II. REGIONAL SETTING : : ; : - : - : 25
III. TREATMENT OF SAMPLES , , ; a “ : 32
IV exruRES 5. : : - : : - : : 34
V. OPTICAL PETROGRAPHY . : é 4 46
(a) Petrographic dassipcation of grain Wipes < ‘ : 46
(b) Dispersal of grain types and vectors. : 7 > 48
(c) Comparative petrography . : - : . : 69
VI. X-Ray MINERALOGY . : : ; : : - 72
(a) Distribution of janerats : : - : : - 72
(b) Phyllosilicate relationships : c : - ; 77
VII. DistRIBUTION OF ELEMENTS . é ; ; - : - 79
VIII. Discussion : : : . : : : - 95
(a) Sediment dispersal 5 : : é : - 96
(b) Distribution of chemical ements : : . 98
(c) Compositional classification of sediments : : : IOI
(da) Textural and mineralogical relationships ; : : 104
(e) Textural and petrographical relationships . : : 108
(f) Relationships between chemistry and texture . : : 108
(g) Tvace elements and gross mineralogy . : é c T1Z
(h) Element-to-element relationships . : F - - 113
(1) Some general geochemical relationships : : : 118
(J) Comments on economic applications . : : . 121
IX. SUMMARY AND CONCLUSIONS . , : : é : : 124
X. ACKNOWLEDGEMENTS . ; : : - - : : 27
XI. REFERENCES ; : : : : - : : - 128
SYNORSIS

Marine sediments in the shallow waters of northern Cardigan Bay, Wales, are dominantly
fine-grained sands that have been well sorted by active tidal and longshore currents. Minor
accumulations of bimodal deposits of sand and gravel are present near the sarns, and at the
base of the eroding coastal cliffs. The bay deposits are composed of six common minerals
(determined by X-ray analysis) : quartz, muscovite, chlorite, orthoclase, plagioclase, and calcite.
Rarely, minor amounts of dolomite are present. Based on their gross mineral content, the
deposits are classified as quartzose sands, with a few being subgreywackes, or sublithic sands.
Petrographic thin section study of impregnated samples, utilizing an inclusive, empirical grain
classification scheme, provides a method for reporting data on, and charting the distribution of,
the several types of quartz and lithic fragments. Charted dispersal patterns indicate local
sources along the coast, namely eroding exposures of glacial debris and sea-cliffs of slates and
greywackes. Spectrochemical data for Al, B, Ba, Ca, Co, Cr, Cu, Fe, Ga, K, Mg, Mn, Na, Ni
Pb, Sc, Sr, Ti, V, and Zr are reported (Tables III—VII), and their distributions in the bay sedi-

MINER. 2, 2. 2

22 SEDIMENTATION IN NORTHERN CARDIGAN BAY

ments charted. Distribution patterns of certain elements associated with accessory minerals
define zones of tidal currents, while elements related to aluminosilicates largely indicate the
locations of eroding coastal exposures currently supplying detritus to the bay. Enrichment
of Ca and Sr is indicative of bioclastic debris. These findings are supported by comparison
of chemical data for nearby Welsh rocks (Table II). It is concluded that most of the bay sands
are derived from coastal erosion sites and not from sources inland; streams are not active
suppliers of sand to the bay ; estuaries are being infilled from the sea ; net transport of sediment
is northward inshore, but seaward, in part, offshore; and that the present bay sands would
have the desirable features of a reservoir rock if they were buried in the subsurface. Finally,
the reported data and descriptions provide useful interpretive criteria for the recognition and
correlation of ancient sediments deposited under similar conditions, as well as for detailed
comparative studies of other Recent sediments.

I. INTRODUCTION

TuHIs report presents the results of a study of 262 sediment samples, collected from
a portion of northern Cardigan Bay, Wales, and 26 nearby reference samples (253
sediment samples BM 1964, 88-340; 26 reference rock samples BM 1964, 341-
366, and thin sections are housed in the sea floor collections of the British Museum
Natural History). The area (Text-figs. 1, 2) is bounded by Sarn Wallog (near
Aberystwyth) on the south and Sarn Badrig on the north. The western boundary
is a north/south line through the West Prong light buoy at the seaward, or terminal
end of Sarn Badrig; thus, the area is approximately 13 nautical miles wide and
23 nautical miles long. The eastern boundary is, of course, the shoreline, including
the Mawddach and Dyfi estuaries (Text-fig. 2). A few samples were collected
beyond the boundary-forming sarns and in several nearby rivers to provide an
overlap with adjacent surveys currently in progress. Sampling profiles and stations
were planned after study of available charts, but the final density of sample coverage
was controlled partly by limitations of boat operation, which is hazardous near the
sarns, and partly by the texture of the bottom as megascopically determined at each
sample station.

This study is a contribution to the Cardigan Bay—Irish Sea Research Project of the
University College of Wales, Aberystwyth. The Project was instituted in 1962 to
provide for a comprehensive, continuing investigation of the sediments, fauna,
oceanography and tectonic history of the southern Irish Sea. The present study
of the sediments of Northern Cardigan Bay is a functional part of the programme
and has the objectives of describing the sediments, in terms of their texture, composi-
tion, chemistry and optical petrography and of determining their distribution and
dispersal. Standard sampling and laboratory techniques which are applicable,
for the most part, to lithified sediments have been used. The results are mainly
presented in a related series of charts, and the conclusions are based wholly upon
the reported data (Tables III—-VII).

Certain aspects of the research, e.g., microfauna, accessory minerals and Pleisto-
cene history are being investigated by others. Although detailed provenance
studies are being conducted at Aberystwyth, it is desirable to report the mineralogical
and spectrochemical data (Tables I and II) for several reference rock samples col-
lected, mainly, from outcrops adjacent to the study area. These are included for

23

SEDIMENTATION IN NORTHERN CARDIGAN BAY

seg

SPA

09
£9

shud

Ta Ted

I €€ I€ Lz
I Le gz ab
cz 9f Le —
—- ol ZI +

I o£ get ice
Z 6z oF cz
ie 6z 9f£ Lz
— gt of g

z ze FE gz
I gz Ib we

20) XO) eo)

a1Ayd
-ouri5

aqqoAyyy
aseqrid,
oyUSIy
UIDA
zyieng
9711010

oyUely
oquery
aqmery

eTeys
oqytuery

opuSry
oyUSIy
oqyuery
oyuery
oyue1y

ayeys
oqUsIy
oyIUeIy
oyuoIy

o}UsIy

ayeys
9U0}S}IS
apyueIy
ayes
ayes

adAj} 3902

ueliquie) UT
SOLIOS
oIseg IOMOT ‘UPIOIAOPIO
ueliquie) UT
AvjD Jeplnog ut ‘Areurszen(d)

AIQAOpuUryyT “T UT
ueliquies) uy

spin ueMmjog ‘ueliquie)
S}IQ souryy ‘ueliquieD
syUQ Soummy ‘uequre)
ajeys osoursuryy ‘ueliquie)
spy yynowieg ‘ueiquie)

sy Y4NoulIeg ‘ueliquie)
sse[q uepUIey ‘ueTIquie)
sSeq PIsIA ‘ueTIquies)
S8eLJ PISIA ‘UeLIquIed
aTeys soyrued ‘ueliquie)

aeys soyiueg ‘uelquies
spog Soruysoy.q ‘uetiquies)
spog ATesjoq ‘uerqureg
AoAOpuryy “JT
SPIED
yqAmysdioqy ‘AraAopuryy ‘9
s}IIH
yqAmysdroqy ‘AraAopuryy “9
AIOAOpUrTT “Q
ArvAopureyy “9
AraAopuryy *Q
Azoaopuryy “9
UWOT}LUIIOF [LITSOTOa4)

(seqzeuoqies seg ‘siedspyey sje ‘ sazyeortsoyjAyd *sAyq
$ ayrmojop “joq { eyoTeo “Teg { aseporsesd “8g { asepoyzI0 “IQ ‘ zy1eNnb 230) ‘ aytI0TYyO “YO ‘ Por “1yN)

(% AVY-X) ADOTVUANIP : SHOOY AONTUAIAY HSTAM

J aTavy

991S0L/HS

zbiSlo/HS
tZ191Z/HS
66618S/NS

6£6499/NS
Lizz€L/WS

ceeblz/HS
9£1£19/HS
ofzgzZ/HS
€z€96S/HS
oL1L19/HS

9z£g6S/HS
9911z9/HS
Z61Sz9/HS
o€zgzL/HS
Lobzg9/HS

g6161Z/HS
bgib69/HS
S91z99/HS
6£6L99/NS

6$gz6S/NS

888909/NS
zS6Lg9/NS
1£0S6Z/HWS
6100¢bL/HS
z96819/NS
‘JOU PUD

99£ ‘ho6r

€9£ ‘bg61
ZOE ‘bo6r
LSE ‘bo61

HSE ‘hob61
zb€ ‘bob

CHE ‘bobr
9SE ‘hQ6r
PE ‘FQ6r
gr€ ‘bo61
gS ‘ho61

LYE ‘boQ61
6SE ‘bo6r
og€ ‘FQ6r
EVE ‘boQ6r
gt ‘ho6r

IFE ‘bo6r
CoE ‘bo6r
tof ‘bg6r
€C€ ‘bo61

IQ€ ‘FQ6r

CCE ‘Fo6r
zS€ ‘bob1
1SE€ ‘Fo61
oS€ ‘hoor
6FE ‘hQ6r
‘ON ‘Wa

SEDIMENTATION IN NORTHERN CARDIGAN BAY

24

bor tir oSg9 LEz S.1F I Z-Z1 «49G-1 «S61 IO-I og-z g-b€ 6g-h gL ghr z-9 11-0 og6z CL ¢.o9€ gg ‘bg6r
ZZI 6S ooze 66 o-Lt 9 9:gI oS-z ofS zg.0 64-0 g.oz tg-b o-F Tie 9:4 gb.o SEE 9 g-F €gE ‘bob1
oS g6r1 oozg zbz L.bE — 0-0g Z1-1 ogL o1-€ 09-0 9-61 gh-6 o.€9 oF€ o.gf o1-g ozI €z 72.1% ZQE ‘bob1
olz Lg oolb 6g g-61 1b 6G.gb go-r ozL 02-1 19-1 0O-be z9-9 0-0S 9g z-9gI o1-0 GLb Zo L.Gt LGE ‘bo61
_ — — oz _- — — — — — — —_ — — — — — ¢ 2.0 PSE ‘bob
gZ ozz oobg SbE S¢.bb Zz o.zh €z-r oSL o€.€ gr.o z-¥z oF-6 0-96 gL o.G€ ¢C.¢ 6x ZI S$ .9% zh€ ‘bo61
I9QI Ott oo9$ I1Ir o-.oF I 0-9£ LE.1 ozh zg-I O1-% 0-02 oOf-g G.zS Ig 0-1z £g-0 oor Gz z.Gr CHE ‘Fo61
o6z Sg oS6€ zg — I 6.br z-r ofo0r gbt.-o 7zS.0 +-.6 oz-b L.z IF 9:6 oO1-0 o61 LEAS) gSE ‘bg6r
gSz =LS ooof Seg 9-71 I 1-br Lo-r o6€ JZ€.0 zr-r Z-11 62-4 F+.6 cE Z-6 €g-0 ozz Iz ©. FE ‘bo6r
zbr gd oSobr £9 o.bz II 0-g9 00-I ootg 06:0 06-1 9g-Iz oz-g o-£g Sg 0-99 £z-0 obb oS z.€1 ght ‘bo6r
o6€ If oSlz +g €.o1 + €.11 gz-1 of6 Lb.o 6b-o L.g oo-F 6.1 1€ 6.b +€.0 ot L 0.9 gSe ‘Fro61
€z1 1Z oozh 98 L.1z I z-Gr +tb.r ogir 99-0 LL.1 g-bi Gg.€ +.bI OF 6-¢ zl.o SSb QI g-II LYE ‘bo61
ofr 68 oS6h +g 0.o0£ II o-1G giI-r ooI1€ go-r 00-2 +-€z 06-9 0-69 gf 0.0€ gi-o oL¢ €b 9g.0z 6SE€ ‘bob
tol 99 oSgf z6 b.Gr z 0-62 gf-1 obL oo-.r Go.r L-b1 OGS-b g-oz +6 L-8 72-0 Sit €z 6.01 og€ ‘bg61
bzI 66 009 601 o-FEt I 0-9$ Sbh.t Sze bz.1 Sg.z 4.62 66.6 0.66 Lit +.St of-0 0624 gf S.1z Eve ‘ho61
OII gir oS1g «lt oz 8 0-Gg Lz.1 GLI o1-I zI-z@ 0-62 GS.L 0.€L tor z-62 gb-o 062 th 0.92 gt€ ‘bo6r
Q9zI 96 oSZ$ of€1 g-SE z 9-09 61-1 oz€ GG.1 GL.1 G.zz 09-g S$-SE ozr L.bz 61-0 OSE Ib o.€z IPE ‘1961
zEI or oSzg O6br o.€€ 61 o-gb GS.r blr 06-0 +9-z 0.0€ Sz-g 9-62 QZI Z-ZI gI-0 O16 £9 ¥+.zz Sof ‘bQ6r
ceg og ogfr 611 S.oF OI €.¢ zz-z o1S II-o z€-€ o-.bz Gz-€ o-1 — — z-o o9f G¢ 7Z-OL bof ‘bo61
o1S 6€ oo6€ 96 4.6 Zz 0-42 gg-0 oo1r €£.0 9£-0 ¥-S 06-.€ F.zI_ I€ o-€1 L1-o ott ge ©. €SE ‘bob1
Z1r €11 ooff tor L.gz + 0-7 g1-I ogg €2-1 gi-z S.1z oLf-L 0-99 ZOI 1-02 £g-0 ogz gg o.Sz 19€ ‘bg6r
oot gtr oooL of1 o.€tr + 0o-G$S zz.-1 ozlL tg.1 Lz-z 9-Qgz o1-g G.9$ 6£€1 o-FI HI-0 O6€ 9g L.G€ CCE ‘bo61
ZII III o0zg g6 o-€€ 0g 0-0$ 11-1 ofg £€6.0 £€g-1 o-Lz obf.L S.gz git £€-0z to-0 olb ¢.Sg €-L1 zG€ ‘bobr
00z 8699 oogt €Z 0.€z £9 0-7 98-0 oogr oz-1 £€6.0 L.z1 Sg-L o.Sb zg z-Sz 7g-.0 OSI 6z S.o1 ISE€ ‘bob
9$t zr ooo9 Lir S.get I o-€L Sz.1 og€r of-1 7zS-.z 9-62 00-g 0-7 96 0-0€ Lo-o oF tL z.bz oS€ ‘ho61
6O€r €11 of1G Seg 9-SE T b.bG Gz.r 096 $€.1 zI-z .gz OS.g8 0-99 £6 G.lz~ z1-0 ot LG Viele OE ‘bQ61
wdd wdd wdd wdd wdd wdd wdd % wdd % % wdd % wdd wdd wdd % wdd wdd % ‘ON ‘WA

IZ A IL aS 9S qd IN ®N UA 38H e) ®O.9f 10 FD OF BF eE a ‘OV
(OIHAVAUDOULOAdS) AULSINAHD { SMOOY AONAUAAAY HSTAM

II 21avy

SEDIMENTATION IN NORTHERN CARDIGAN BAY 25

chemical comparisons, and are not intended to represent, or to suggest, a basis for
establishing overall provenance-sediment relationships. Such is beyond the scope
of the present study.

There are no published results of Recent sediment investigations for this area.
Blundell, King & Wilson (1964, p. 47) report that, on the basis of seismic evidence,
some rocks beneath Cardigan Bay may be “a sequence of late Mesozoic or Tertiary
clays and sands’”’. Their research has been confined to geophysical investigations,
and only indirectly relates to the surface sands. Haynes (1964) made brief mention
of the sand texture for this area. Adams, Haynes & Walker (1965) reported boron
data for Dyfi estuary sediments.

i REGIONAL SETTING

Although the present study is restricted to a limited inshore area in northern
Cardigan Bay (Text-fig. 1), the influences of bathymetry, physical oceanography and
coastal geology beyond it establish, in part, the regional picture and, thus, are
discussed below as are the local environmental influences.

Bathymetry. Text-fig. 1 shows that the sea bottom slopes away from the Welsh
coast until it reaches a depth of slightly over 50 fathoms (300 feet) at the bottom
of St. Georges Channel. The depth contours are approximately parallel to both
the coast and to the north/south axial salient of St. Georges Channel itself. The
overall gradient between the Welsh coast and the 20 fathom line, a contour delimiting
the seaward margin of Cardigan Bay (H.O. Publication 145, 1951), is some 4°5 feet
per nautical mile. The only notable gradients in Cardigan Bay proper are the
sharp drop-off to about five fathoms immediately adjacent to parts of the coast,
usually within one mile of the shore; the relatively steep gradient surrounding the
sarn ends, and finally the entrenched trough, defined by the 20 fathom line, extend-
ing northeastward into Tremadoc Bay. The “ trawling ground ’’, a slightly deeper
(12 fathoms) depression south of Aberystwyth, is locally important as a site of mixed
clay and sand deposition.

Beyond the 20 fathom line, in the IrishSea proper, the bottom slopeshows a slightly
increased gradient between 30 and 40 fathoms, but otherwise it remains fairly
constant seaward. The deepest part of the Irish Sea (50 fathoms) in this region
forms a narrow north-trending depression from 5 to 15 nautical miles in width.

In directing attention to the bathymetric features of the study area itself (Text-fig.
2), it is seen that the only pronounced features present are the three sarns. If one
neglects their local influence on bottom topography, the inshore province along the
coast of Wales, between the sarns, is one of gradual increase of depth seaward.
Nevertheless three minor features of the bottom should be mentioned. The first
is the depression at the landward end of Sarn Badrig. This depression is scoured by
the strong, nearshore current which is initiated by overflow (at flood tide) from
Tremadoc Bay. It is important as an active transport passage for detritus from
Tremadoc Bay. The second is the shoaling zone which is in alignment with the
lobate contours extending from the northeast corner of this survey to near the
seaward termination of Sarn Bwch. On the basis of these bathymetric contours

26 SEDIMENTATION IN NORTHERN CARDIGAN BAY

alone, one may visualize a definite zone of active sedimentation extending from the
northeast corner of the survey to some several kilometres west of Sarn Bwch. The
third is an open trough developed at the outer end of Sarn Bwch; it is delimited
by the ten fathom line. This bathymetric depression may explain the relatively

STremadoc.
wd i Bay

Fic. 1. Chart showing location of the study area in relation to the principal geographic
and bathymetric features of the southern Irish Sea area. Contours in fathoms.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 27

high velocity tidal current which is encountered off the end of Sarn Bwch, for
although of modest dimension, it may force considerable water into a relatively
narrow passage along the bottom there.

4

KILOMETRES

e215
West Prong

2250

Machynileth
65 60

2276 +67 ,66

\'70

+199
Gh -198

a joy ge SARNS
= ee A BADRIG —BA
Orc Bate re BWCH —BW
: WALLOG — W

SES

Fic. 2. Location of samples taken from Cardigan Bay and the local rivers.

28 SEDIMENTATION IN NORTHERN CARDIGAN BAY

Climatic factors. Of the various climatic factors, only seasonal gales, rainfall and
temperature are of consequence in the present study. Gales with a velocity of
38 miles per hour and upwards, are common over the Irish Sea and are known to
be the cause of heavy seas along the Welsh coast. The average number of gales
(H.O. Pub. 145, 1951, p. 14) in this area was 35-3 per year for the years 1876/1915,
of which 27-7 occurred during the months of October to March and 7-6 during the
period of April to September. Of the 35-3 gales per year, 14-5 were southwesterlies,
and 10-3 were northwesterlies. Winds from other directions may occasionally reach
gale force, but they are short-lived, less common and are considered unimportant
in this study. The importance of the westerly, winter gales cannot be over-empha-
sized, because the erosive action of the storm breakers causes most of the coastline
modification and subsequently furnishes the bulk of the detritus now entering the
bay.

The region, as a whole, experiences moderately heavy rainfall, e.g., the annual
rainfall average for the years 1881/1915 at Aberystwyth was 46 inches (H.O. Pub.
195, 1951). As the topography of the land controls, to some extent, the rainfall
from the moist wind off the Atlantic, the mountainous Welsh coast experiences a
higher proportion of rain than those areas of modest relief nearby. Heavy rainfall
along the Welsh coast accounts for the lower-than-normal salinities encountered
within the first 2 or 3 miles of the coast (Dr. John Haynes, personal communication).

Coastal features between Aberystwyth and Mochras. Greywackes are exposed
along the shore between Aberystwyth and Borth (PI. 2, fig. 1), and have been studied
in detail (Wood & Smith, 1959), particularly their sedimentary structures. These
authors report that, on the basis of thin section study, the greywackes are poorly
sorted, i.e., containing a considerable amount of fine material with the grit, and
contain quartz (75 to 80% of the material which could be resolved in thin section),
feldspar, muscovite and carbonate minerals, mostly in sand and fine sand sizes.
The cliffed shoreline between Aberystwyth and Borth is being actively eroded, and
the lithic fragments occurring at the base of the cliffs are predominantly composed
of, and immediately derived from, the Aberystwyth Grits.

A storm-beach, composed of pebbles and sand with a beach of medium-grained
sand, extends from Borth northward for some three miles where it terminates at the
Dyfi Channel. Behind the Borth beach an estuarine development of marsh and
open sand flats extends up the Afon Dyfi Channel from the river mouth. These
sediments texturally resemble those offshore, and grass, in part, has aided in stabiliz-
ing this major accumulation of sand.

From the town of Aberdyfi northward towards Towyn, the coastline consists of a
moderately wide beach backed, in part, by sand dunes. The coastal section here
is aggrading, and sand dunes behind the beach have become stabilized and grass
covered. In the vicinity of Towyn, however, the coast is being actively eroded.
Outcrops along the coast in this area are mixed gravel and sand beds, boulder clays
and peat deposits. As these relatively unconsolidated deposits are easily eroded by
high seas, erosion of the coast at Towyn is proceeding rapidly, and this site is an
important source for much of the detritus now being delivered to the bay. A

SEDIMENTATION IN NORTHERN CARDIGAN BAY 29

coastal section of boulders commences near Towyn, in the vicinity of Sarn Bwch,
and extends northward, in part, around the point at Tal-y-garreg. A small portion
of the coast northwest of Tal-y-garreg is exposed bedrock of slates and thin, dense
sandstones. A fresh exposure of boulder clay, at the head of a boulder beach,
outcrops about a mile north of Sarn Bwch point (PI. 2, fig. 2). The boulders eroded
from the glacial debris in this area have formed a natural rip-rap, which partially
inhibits erosion of the remaining boulder clay and, in a few places, vegetation is
growing over the boulder clay. Nevertheless, field evidence suggests that some
erosion is still taking place—perhaps modified by the protection Sarn Bwch affords
from southwesterly seas—and that the finer clastics are being delivered to the sea.
The coast between Ty-wen and Station 134 is reasonably stable, as suggested by
field evidence. From Station 134, south of the town of Barmouth, a long storm-
beach of mixed sand and pebbles extends northward to the Mawddach estuary
mouth, where it curves inward and parallels the channel.

Much the same setting is found for the Mawddach estuary as that found for the
Dyfi, i.e., behind the protective storm-beach, an area of fine sand, partly covered
by Spartina grass, extends up the estuary. In the upper estuary, local areas of
marsh deposition occur on both sides of the channel, and, in places amongst the
marsh sands, argillaceous silt is found as a very thin surface deposit.

From the town of Barmouth, a relatively wide beach extends northward continuing
to, and forming, the outer edge of Mochras point. For the most part, this coastal
zone is one of sand accumulation resulting from northward transport accompanying
inshore tidal currents and wave induced, littoral drift.

Field study of the river deposits did not reveal active transport of fine sand.
The transport of fine sand, under normal river flow, is apparently negligible at the
present time. To say that the rivers may have once carried an abundance of fine
sands cannot be established on the basis of present field evidence, and, thus, the
past sediment transport history of Welsh rivers is unknown. On the other hand,
it is probable that clay is transported by the streams, in flood, possibly in the form
of muscovite and chlorite flakes, but the high energy regime of this part of Cardigan
Bay prohibits its deposition.

Local geology. The general geologic structure and rock types of the area adjacent
to northern Cardigan Bay have been described by Keeping (1881), Jones (1912,
1956), Jones & Pugh (1935), Cox (1925), Cox & Wells (1920, 1927), Jehu (1926),
Challinor (1949), Wood & Smith (1959), Mohr (1959), and Matley & Wilson (1946).
Additional studies are listed by Thomas (1896) and Bassett (1961). The majority
of the exposed rocks in the area are slates, shales and arenites, all of which are folded
and faulted to varying degrees, and in many places associated with hydrothermal
(vein) quartz. In a few places, crystalline rocks are exposed, the most important
types being granophyres, dolorites, rhyolites, and several other volcanic varieties,
but by far the greatest number of exposures are those of slates and greywackes.
Geochemical and mineralogical data for 26 of these rocks are given in Tables I and II.

Boulder clay covers much of the countryside along the coast, but with the excep-
tion of coastal exposures at Towyn and north of Sarn Bweh, it is not being extensively

30 SEDIMENTATION IN NORTHERN CARDIGAN BAY

eroded and, indeed, is usually grass covered. Pleistocene and subsequent trans-
gressive reworking is undoubtedly important to the Irish Sea sediments as a whole,
particularly in the distribution of coarse clastics, but until more is learned of the
mineralogical and chemical nature of the boulder clays, and their relative distribution
and stratigraphy, and until long bottom cores are studied, no realistic historical
interpretations can be made.

Salimty and Temperature. Salinity determinations have been made on water
samples collected seasonally both from surface estuarine waters, and from surface
and subsurface waters offshore by Dr. John Haynes. His investigation has shown
that salinity varies in the Dyfi Estuary between 1%, and just over 36%, whilst
inshore waters, of northern Cardigan Bay vary in salinity between 29%, and 34%, ;
the former value being typical of the mixed waters just seaward of the estuary
mouths during periods of ebb flow. Data for “typical”’ offshore waters, at least
seaward for some I5 miles, show a gradual but definite increase in salinity up to
34%. This figure is lower than open Atlantic salinities of about 35%,, and is the
result of fresh water entering the sea through the numerous rivers along the Welsh
coast. While salinity is primarily of biological importance, it may influence some
trace element concentrations in detrital marine clays, particularly boron (Walker
& Price, 1963), and perhaps in the phyllosilicates in lithic fragments as well.

Temperature data for part of the outer Cardigan Bay region are reported by Lee
(r960) in his recent study of the Irish Sea. Some serially collected data have been
acquired by the Aberystwyth staff, but only for the estuarine and inshore waters
in the southern part of the present study area. From a series of surface measure-
ments in March 1953, the 6-5° C. isotherm was found to extend in a north/south
direction some 17 km. off the Welsh coast, and very near Station 118 of the present
survey. The 7°C. isotherm parallels the 6:5° C. isotherm about 30 km. offshore,
and the highest temperatures (7-5° C.) were recorded in the middle of the Irish Sea.
Temperatures on the Irish side were somewhat warmer than for the Welsh side,
70° C. being an average (Lee, 1960, p. 17). During the autumn of 1962, the author
and Dr. John Haynes measured sea surface temperatures for the area between Sarn
Wallog and Sarn Bwch. Inthe month of October a high value of 14° C. was recorded,
which decreased to 7° C. in early December. Littoral water measurements were
made at Aberystwyth throughout the severe winter of 1962-63, and the lowest
temperature recorded was 0-7° C.

Waves and currents. In general, the effective fetch for winds from the west is
about 80 to go nautical miles. For southwesterlies, considerably more fetch is
available as the opening of St. Georges Channel is in line with the open Atlantic.
In this instance, effective fetch may exceed 200 nautical miles, although refraction
of long-period waves by St. David’s Head, Pembrokeshire, diminishes the direct line
of wave parallelism into Cardigan Bay. For northwesterly winds approaching from
greater than about 300° (true), the effective fetch is greatly shortened by the Lleyn
Peninsula and even more so by Sarn Badrig. Thus, prevailing winds from the west
and from the southwest, with longer fetches, initiate the highest waves encountered
along the coast of northern Cardigan Bay. Inasmuch as the prevailing sea, as well

SEDIMENTATION IN NORTHERN CARDIGAN BAY 31

as the gale force winds, is from the west and southwest, most of the waves striking
the shore provide a net transport of nearshore water in a northerly direction and
immediately adjacent to the coast. In the case of northwesterly winds, the effective
fetch, as previously mentioned, is only some few tens of miles and, as such, wind
waves striking the coast from this sector may initiate weak, wave-induced littoral
drift to the south. Sarn Badrig acts as a natural wave trap for northwesterly wind
waves coming across Tremadoc Bay; thus, only those waves with a reasonably
short period or relatively small orbital geometry will finally break on the shore
between Sarn Bwch and the vicinity of Barmouth. Likewise, the combined effect
of the Lleyn Peninsula, Sarn Badrig and Sarn Bwch strongly diminishes the size
of waves approaching from the northwest. Superimposed upon the westerly
wave trains is a complex pattern of wave refraction caused by the pronounced
submarine topography of the three sarns.

Considering this study area from an environment and energy viewpoint, the sarns
probably provide natural wave traps for all reasonably long-period waves striking
the coast obliquely. Bottom sediments in the lee of oncoming wave trains are not
wave agitated below a depth controlled locally by the height of water over the
sarn. This assumption is supported by textural evidence, for the deposits offshore
from Borth are some of the finest grained clastics found in the area. The problem
of interpreting waves and their refraction patterns in the light of sediment transport
mechanics is made extremely difficult by the complex tidal currents. In general,
however, three observations may be made regarding the energy system in northern
Cardigan Bay. First, the net transport of nearshore detritus is to the north, but
detritus may locally be transported southward for short periods of time when the
prevailing wind is from the northwest and the tidal current is ebbing. Second,
the fetch presented by the entire width of the Irish Sea is such that gale force winds
cause a heavy surf along the coast, even though refraction by the sarns reduces or
spreads out the available energy. Third, the area is a “ high energy environment ”’
with strong tidal currents, drift and surf.

The only detailed information available on the direction and magnitude of tidal
currents in the northern part of Cardigan Bay is that published by the Admiralty
Hydrographic Dept. (1962), and observations, descriptive in nature, made at
Aberystwyth. The latter were restricted to the estuary mouths and to the three
sarns.

The major axis of offshore tidal currents, both at flood and at ebb, is a north/south
one for the region between Sarn Wallog and Sarn Bwch. In the area between
Sarn Bwch and Barmouth, the axis of the tidal currents parallels the coast and is
oriented in a northeast /southwest direction. However, tidal current vectors on
H.O. Chart No. 4454 (1934) show a tidal current direction almost normal to Sarn
Badrig on the flood tide, which would make it northwesterly, but a due south
orientation on the ebb tide. This shift in direction for the two tidal phases presents
complications in developing a rigid interpretation of currents for the northern area
of the survey, and until adequate measurements are made at sea, preferably from
an anchored vessel, all conclusions must be considered as provisional ones. Fine
grained sand seems to be transported over the sarns at shoaling depths, since it is

32 SEDIMENTATION IN NORTHERN CARDIGAN BAY

entrapped among the seaweed attached to boulders and cobbles along the axis of
Sarn Wallog and Sarn Badrig. The presence of this sand is not surprising as strong
currents were observed in the immediate vicinity of each sarn, in fact, some currents
accelerated from 3 to 5 knots over the crest of the sarns. These submarine topo-
graphic features restrict the flow so that the cross sectional area normal to the
transport direction is diminished, and there is a concomitant increase in the velocity
of the current. Inasmuch as the surface velocities for the north/south tidal currents
have a mean of almost one knot at maximum flood, it is not unreasonable to estimate
three to fivefold increases in velocity over bathymetric highs.

One area in particular should be discussed, for it shows a major tidal current
influence on the dispersal of sands. This is in the right angle corner formed by the
axis of Sarn Badrig and the Barmouth/Mochras coast. A depression found very
near the coast is suggestive of a zone of strong bottom current scour and inshore
sediment transport. Admittedly, coastal drift in this area provides for the north-
ward transport of sand, however, ebb tide and possibly overflow, caused by water
pushed into Tremadoc Bay by southwesterly gales, creates a strong net transport
nearshore and over the sarn in a southerly direction. Fine sand is, in fact, trans-
ported over the sarn and to the south as shown by the dispersal patterns for several
grain types.

According to Hjulstrom’s (1939) curve the tidal currents in the northern part of
Cardigan Bay are of sufficient strength to remove clay size and fine silt clastics from
the area, and, there is in general, little opportunity for clay detritus to accumulate
in abundance with the offshore sands, especially since no major depressions are
present within this area.

Itt. TREAGMENT OF SAMPLES

In the laboratory surplus water was removed by decantation or by a rubber
syringe. The moist sample was then placed in a drying oven for a period of approxi-
mately 48 hours at a temperature of 100° F., or slightly below. Every effort was
made in handling the samples to avoid removing the constituent sea water which
adhered to the grains. Sedimentologists, notably Goldberg & Arrhenius (1958),
have suggested that the constituent, or interstitial, water must be considered as a
part of the lithosphere rather than as a part of the hydrosphere. This is particularly
important concerning those trace elements which may occupy grain surface sites.

A representative portion of between 80 and 100 grams of sediment was spread on
a sheet of paper and all unbroken shells above 4 mm. in size were removed. Small
shell fragments were considered a constituent part of the deposit. A portion was
weighed and then passed through a set of British Standard sieves on a Ro-Tap
shaker. Using both 3-cycle semi-logarithmic and arithmetical probability paper
(Chartwell Nos. 5535 and 5571), cumulative curves were drawn for all analyzed
samples (Milner, 1962; Folk, 1961). Gravel and cobble samples were measured
using machinist’s calipers, weighed and the resulting data graphed to provide
cumulative curves (Krumbein & Pettijohn, 1938).

In order to provide a basis by which Cardigan Bay samples could be petrographi-
cally compared and to delineate dispersal zones in the bay, the writer studied thin

SEDIMENTATION IN NORTHERN CARDIGAN BAY 33

sections of the various samples, excepting some gravels. The procedure for preparing
thin sections of Recent sediments has been published elsewhere (Moore & Garraway,
1963). At least 330 grains were identified in each thin section (van der Plas, 1962) ;
while for some slides more than 500 grains were counted (Weber & Middleton, 1961,
Pp. 247).

The sample portion held aside for X-ray diffraction was further divided into three
parts using a small Jones splitter. This material was then ground for 25 minutes
in a power driven agate mortar and pestle, and sieved through a 325 mesh A.S.T.M.
sieve. Approximately 0-5 gram of the sieved powder was back loaded, i.e., with the
“face ”’ side against a clean, smooth glass slide, into a standard Phillips aluminium
sample holder. Three mounts were made of each sample.

A Phillips N.A. wide-range, Geiger counter, X-ray diffraction unit and a Brown
“ Electronik ’’, fast response, chart recorder were employed. Operating conditions
were as follows : copper target, 40 kV, 17 MA, Ni filter, 1° diverging and converging
slits, scanning (goniometer) speeds of 1° and 2° minute, recorder chart speed of
I inch/4 minutes, and the rate meter set at a scale factor of 8, multiplier at unity,
and time constant at 4 seconds. For mineral identification, one mount was analysed
at a goniometer speed of 1°/minute from 4° to 66° 26. The diffraction maxima or
peak positions were read to the nearest 0-01° in the range from 4° to 34°20. Mineral
identification was made by comparing the unknown (sample) diffractogram with
known diffractograms of previously analysed mineral standards, and by comparison
with the d spacings listed in the American Society of Testing Materials X-ray Powder
Data File. Quantitative mineral data were obtained by subjecting each of the three
mounts to diffraction analysis from 4° to 36° 2 6 at a goniometer speed of 2°/minute.
The characteristic diffraction maxima of each of the minerals were measured in
tenths of an inch above the background, as recorded, and the mean taken for the
series. These data were then multiplied by mineral intensity percentage factors
previously obtained from the standards analyses and, finally, normalized to total
100% of the minerals identified.

Compositional mineral abundances obtained in the above manner very closely
approximate the true or actual mineral percentages. The difference is due to
variations in the degree of crystallinity, and to slight variations in the mass absorp-
tion coefficients between the standards and the sample material. Furthermore,
variation can result from preferred particle orientation, although the latter can be
minimized by mounting procedures. In order to avoid confusion between 14 A
chlorites and the vermiculite and montmorillonite groups, both heat treatment and
addition of vapourized ethylene glycol were standard procedures. For comparative
purposes, reference rock samples containing montmorillonite (from the Isle of Wight)
were subjected to diffraction analysis along with the bay samples.

The portion of each sample retained for analysis by emission spectroscopy was
ground to the fineness of a silt-clay powder, using a Braun 6-S grinder fitted with
Coors Co. “ Alumina” brand grinding discs. A small portion, about one gram,
was then transferred to an agate mortar and hand ground to powder fineness. This
portion was dried at 110°C. for 24 hours. Duplicate fractions of 25 mg. portions
were packed in United Carbon Co. type 3417 anode preforms. The cathodes were

34 SEDIMENTATION IN NORTHERN CARDIGAN BAY

specially fabricated from }-inch diameter graphite rod, and were purified by refluxing
with dilute aqua regia for three days and with distilled water for two days. The
loaded preforms were stored in plastic holder-boxes within a vacuum dessicator
until arced.

All samples were prepared in the above manner and were subjected to a total
energy burn (arcing) in a Jarrel-Ash Co. Model 7100, 3-4 metre Ebert emission
spectrograph with a dispersion of 5 A/mm. in the first order. Operational adjust-
ment was set at Ig amps, 220 volts d.c. when short circuited across the electrodes.
Eastman Kodak III-o spectrograph plates were used and developed with Eastman
Kodak D-19 developer. Comparisons were made with W-1, G-1, and with special
standards previously developed by W. L. Hall of the Texaco Inc. Laboratory ex-
pressly for sedimentary geochemical interpretations. Pd and In were used as in-
ternal standards and a standard step-filter system permitted determinations to be
made of selected major elements. All readings were made on the same machine,
a console Jarrell-Ash Co. model 2100 microphotometer, and no valves or other
component parts were changed during the reading of the plates.

[Ve TEXT URE

In addition to “ grain size”’, statistical measures provide a comparative basis
for detecting subtle textural variations within a single class interval, relating compo-
sition to texture, establishing texturally controlled geochemical associations, defining
the degree of size dispersion, or sediment type mixing, and charting textural trends
related to environmental energy and to provenance influence. Textural definition
is nevertheless not an end in itself.

Statistical measures. Based upon data obtained from sieve analysis, the cumulative
curve for each sample was drawn and analysed as suggested by Folk (1961). The
cumulative curve is the best graphic form from which all statistical measures are
obtained, and it is considered as the beginning point in modern statistical descrip-
tion. Those measures selected by the author for use in this study are as follows:
first quartile (Q,), third quartile (Q;), median diameter (Md), sorting (So), skewness
(Sk), 16 percentile intercept (¢ 16), 84 percentile intercept (¢ 84), median in phi
units (¢ M or ¢ 50), graphic mean (¢ Mz), graphic standard deviation (¢o g), and
standard deviation interval mean (# MI). These terms, their meaning and deriva-
tion, are reviewed in detail by Folk (1961).

Textural classification terms. Using the statistical measures of ¢ Mz and the first
and third quartiles (Q, and Q;), equivalent Wentworth (1922) and Niggli (Pettijohn,
1957) nomenclature have been determined for each sample. The terms have been
used to prepare charts depicting the textural facies and to provide textural classi-
fications (Text-figs. 3, 4, 5). The restricted two-component nature of the bay
deposits, namely sands and gravels, eliminated the effective use of such ternary
component classifications as were reviewed by Shepard (1954).

Grain size distribution based on 6 Mz measures. Using the reported measure
@ Mz, a contoured chart was drawn for the Cardigan Bay deposits (Text-fig. 3).

SEDIMENTATION IN NORTHERN CARDIGAN BAY 35

eM

Od 2st Als: iG) 78S) Ont oy be
s——s 7 ——t te Vv
KILOMETRES YY fey
s

= ° 1264
. ° B

Barmouth 2
hi @

. L
° SAE

fi | Tal -y-garfeg
\ 32 fa! é “¢

~-- Aiea at ee
a ay “ a a Pulte as \y
“oe AES WA
Se FAN FT «Ng Aytown
Saye

145
Borth
on Le,;
SARNS
BA — BADRIG
BW- BWCH
WA — WALLOG
Cafe Aberystwyth
a Ue efis ON

Fic. 3. Textural chart based on graphic mean size determinations (¢ Mz). Contour
interval r¢.

MINER. 2, 2. 3

36 SEDIMENTATION IN NORTHERN CARDIGAN BAY

The contour interval of one phi unit was purposely selected in order to conform with
Wentworth grade scale limits. The most noticeable area of single-class size distri-
bution is the major region of fine sand (¢ 2 to ¢ 3), which, in this part of Cardigan
Bay, is located inshore and along the coast. This deposit is remarkably uniform in
texture, averaging about 2:4 ¢ Mz, and encloses only two sites of coarser sediments:
a deposit of lag gravel in the vicinity of Station 4 and a small patch of medium sand
(@ 1-8, @ Ig) in the area around Stations 218 and 242. The only deposit of very
fine sand in the bay area proper is found within the major sand body in the area
around Station 25, about 1-5 km. off the Borth beach, an area partially protected
by Sarn Wallog.

Fine sand extends just north of Sarn Wallog across the shoaling area to the south
side of Sarn Bwch. The northward extension of this sand is from the vicinity of
Station 37 completely across the, more-or-less, flat shelf bottom off Barmouth to
the base of the steep, south side of Sarn Badrig. It also extends approximately
15 km. seaward to near Station 14, and as a southwesterly zone through Stations
116, 275, and 193. In the northern area the tongue extends seaward through
Station 246. The fine sand on both sides of Sarn Bwch, which separates Stations
35 and 37, appears to come within a few yards of the ridge of the sarn at this point,
and, in fact, fine sand is transported over the sarn in this area. Safety precautions
during boat operations limited sampling in the immediate sarn area, and conse-
quently the actual width of the connection cannot be established. Petrographic
observations suggest that the fine sand is compositionally similar on both sides of
Sarn Bwch.

Other than the major deposits of gravel and cobbles constituting the sarn ridges
themselves, the only coarse deposits in the area are the tongue of coarse debris
extending northward from Sarn Wallog, two similar coarse deposits extending
northward and westward from the end of Sarn Bwch, and the nearshore and littoral
deposits north-west of Tal-y-garreg, and along the base of the cliffs south of Borth.
(High beach and storm-beach gravels were not sampled for this study.)

The estuaries are largely filled with fine and very fine sands, and rarely silt: the
latter type being restricted to two sites (Stations 124 and 126) on the north side of
the Mawddach estuary, and one small area (Station 145) where the Afon Leri enters
the Dyfi estuary.

Descriptive textural facies. As an alternative presentation of the basic ¢ Mz
measures, a chart was drawn using descriptive nomenclature equivalent to the ¢ Mz
values (Text-fig. 4). The designated Wentworth term for the ¢ Mz measure of
each sample was plotted and all similar, adjacent textural types were grouped as
contiguous sedimentary facies. This system permits a somewhat broader inter-
pretation as sharp changes in facies may be charted without adhering to rigidly
controlled numerically based contour trends. From a study of the chart, it is seen
that the major fine sand area is represented much the same as when using ¢ Mz
(numerical) measures, so are the gravel zones associated with the sarns. The statisti-
cally designated medium and coarse sands, however, are interpreted somewhat
more smoothly, i.e., they exhibit better continuity and more natural distribution.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 37

If one accepts the implications of Hjulstrom’s (1939) curve for detrital transport
and sedimentation, there is no reason why some textural classes should not be
distributed without intermediate sizes between them. Thus, the deposition of
medium sand on the north side of Sarn Wallog and the deposition of fine sand on
the south side do not seem unrealistic cases at all. In fact, the interpretative, but
not necessarily statistical, pattern of depositional facies is more clearly brought out
in this chart based on descriptive grouping. In the broadest sense, that of establish-
ing the major clastic facies, the patterns based on the descriptive chart (Text-fig. 4)
agree with those outlined on the ¢ Mz chart (Text-fig. 3).

Distribution of depositional types based on Niggl terms. In order that mixed, or
poorly sorted, textural types may be charted and compared with others, a chart
based on the Niggli nomenclature (Text-fig. 5) is presented. The Niggli scheme
uses the equivalent terminology of quartiles at the Q, and Q, points (Pettijohn,
1957, p. 26). A binomial designation is obtained when the quartile intercepts fall
in separate, widely spaced classes. For extremely bimodal deposits the scheme is
desirable ; however, for essentially arenaceous deposits (a single Niggli class), the
scheme is not definitive.

From study of the Niggli chart (Text-fig. 5) it is apparent that for sand sizes,
resolution is not possible. In other words, the specific texture of the sands is not
discernible, and the entire area of fine, medium and coarse sands is charted as one
large sand facies extending from the shore to the westward limit of study. On the
other hand, deposits associated with the sarns and with littoral deposition between
Aberystwyth and Borth exhibit distinct class bimodality and are the only bay
deposits which are so defined by the Niggli system. The textural character of
Cardigan Bay deposits is (expressed in its broadest terms) a major sand area broken
only by the gravels and sandy gravels of the sarns, and the previously noted tongues
of coarse sediment. By using the Niggli scheme as a basis for establishing the
textural framework, we are able to illustrate the dominantly sandy nature of the
Cardigan Bay deposits in a very striking manner, and also show that the coarse
sandy sediments are, in fact, largely mixtures of sand and fine gravel. The three
classification systems (¢ Mz, Wentworth and Niggli) provide mutually compatible
results, namely, the bottom sediments of Cardigan Bay are essentially fine sands
with some development of gravelly deposits associated with the sarns and with the
eroded cliff area, and some very limited development of accretionary silty sediments
in the estuaries. The Niggli system has, however, the added feature of showing the
locations of bimodal deposits.

Importance of the size classes. The finest sand sizes (2:5 to 3:0 ¢) are distributed
mainly on the inshore shallow bottom, within approximately 10 km. of the coast,
and in the estuaries. In the northern part of the area, fine sands (2:5 to 3:0 ¢)
are found in the region off Eglwys Llanaber, and just off the estuary mouth at
Barmouth. These findings coupled with the charted medium and coarser sands,
suggest that for this area along the Welsh coast there is, in general, increasing grain
size with increasing distance from the shore. This finding, particularly in relation

38

SEDIMENTATION IN NORTHERN CARDIGAN BAY

2 ALE ASE 4 ee5 § 7aas! Aten:
=i
KILOMETRES

= Barmouth ‘
; \/

SARNS
BA — BADRIG
BW — BWCH
WA — WALLOG

Fic. 4. Chart of grouped textural facies (Wentworth, based on ¢ Mz measurements).
Fine sands are shown by horizontal lines, medium sands by vertical lines, coarse sands
by CS, very coarse sands by VCS, granules by GN, gravel by GR, very fine sands by VFS,
and silt by ST.

SEDIMENTATION IN NORTHERN CARDIGAN BAY

OMim2erS Al on iGi 72 Bem Ont
——— ee td
KILOMETRES

SANDY
= \ GRANULES

ie
Se tS
<a a E .

aN 7

COBBLE) °
GRY. D,

-

-TOBBLES~~*
ae Fines <img
aie e e

GRUNULES

> Eglwys Llanaber

sig

*\ Barmouth

Aberdyfi
oS iS cale

0 De:
ann Fine GRAVEL oe SARNS

; a if Ge ~<a
aa ~eaepue SRY. Ee z ais
WA — WALLOG

Fic. 5. Chart of sediment texture using Niggli nomenclature.

39

40 SEDIMENTATION IN NORTHERN CARDIGAN BAY

to contemporary studies of other Recent sediments, is significant (Shepard, Phleger
& van Andel, 1960).

Predominance of the ¢ 2-¢ 3 interval (0-12 to 0-25 mm.) grains in the bay suggests
that a reasonably stable current energy and sediment dispersal system prevails.
The ¢ 2 class of fine sand has been found as a common constituent of shallow Recent
marine deposits in several other parts of the world, and it may, in fact, be the best
adjusted (to physical environment) textural class of all inshore Recent marine sands,
cf. Shepard et al. (1960). Stride (1963) reports that the sea level (and presumably
the energy relationship) for the United Kingdom has been relatively stable for the
past 6,000 years. This length of time should be sufficient for establishing normal
dispersal pattern relationships between the tidal currents and the available detritus
in the inshore area of the shelf, including Cardigan Bay. Therefore, a balance
between source, tidal current energy, texture, and dispersal of the fine sands probably
exists in Cardigan Bay.

Sorting of the deposits. Sorting, or “ dispersion ”’, is a statistically derived measure,
and need not necessarily be related to the textural size measures. However, in
the study of inshore Recent marine sediments, sorting has been found to vary in
accordance with grain size (Hough, 1942; Krumbein & Aberdeen, 1937; Stetson,
1938). This is also true for the Cardigan Bay deposits. From Text-fig. 6, it is
clear that there are two major sediment types based on their characteristic So values ;
the large area of fine sand between the sarns, which possesses So values below 1:25
and, as such, exhibits excellent sorting, and the coarser grained mixed sands and
gravel deposits associated with the sarns and coastal erosion areas. Some of the
fine sand at stations in the inshore waters just off Borth exhibit truly exceptional
sorting (So): e.g., Stations 7 (I-06), 10 (1-08), 17A (1-08), 18 (1-09), 28 (1-09) and
68 (r-09). These stations are all within the finer ranges of the ¢ 2 to ¢ 3 deposits
found there, and suggest a zone of sediments very well adjusted to the present-day
current-energy system.

One of the limitations of any contoured chart is that the contacts between major
facies are suggested to be transitional. This is not always substantiated in the field,
and it is probable that the change from well sorted sediments to an adjacent area
of poorly sorted sediments is an abrupt one. This became apparent during review
of the data, as very few stations had sorting indices (So) between about 1-40 and
2:00. Reference to the excellent sorting trends in the fine sands and the poor sorting
in the coarser sands and gravels, further established this facies relationship. Fine
sands with low So values, such as Stations 6 (1:20), 207 (1:13) and 16 (1-24) parallel
the rocky coast just south of Borth, and are only one km. seaward from a coastal
zone of mixed gravel, cobbles and sand. The facies contact there is sharply defined,
based on charted statistical measures as well as Marconi echograms. The north-
ward-extending, curving-shoreward zone of poorly sorted deposits (Stations Igo,
IQI, 192, 262, 201, I17, 202 and 20) just north of Sarn Wallog, is surrounded entirely
by fine sands with excellent sorting values, all below 1-25 So. The contact between
these two types is sharply defined, although the environmental mixing of sizes is
obvious.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 41

ORM ica 4 tS G7 SS IO nt
(—— = = Ss Se Sa Se Ss
KILOMETRES

Xa “ 6 ce
Barmouth “of. .

167

“ e CY . \Aberdyf
15) TG,
146 CL 35° °

268 oe ° e
° Aan (

SARNS
ae BA — BADRIG
5.6 BW-BWCH
pe WA — WALLOG

4 Aberystwyth

Fic. 6. Charted sorting values (Trask So measures).

42 SEDIMENTATION IN NORTHERN CARDIGAN BAY

Pronounced So gradients occur in the area north of the terminus of Sarn Bwch.
Stations 228 (5-04) and I00 (1-22) are, in fact, only a fraction of a km. apart. The
transition from well sorted fine sands to poorly sorted coarse deposits approaching
Sarn Badrig is particularly sharp, and on one sampling profile (Stations 217-221),
the sandy bottom was found to extend right up to the base of the sarn cobbles
and gravel; this observation is based on echograms and on check samples taken as
close to the sarn as the vessel could safely be manoeuvred.

With the exception of Stations 146 (1-75) and 167 (5:59), all samples from the two
estuaries were found to be well sorted and predominantly fine sands. Indeed, all
estuary values are below 1-20 So, including two important Stations: 163 (I-12)
at the head of the Dyfi estuary and 139 (1-08) at the head of the Mawddach estuary.
The upstream river deposits, on the other hand, were only fair to poorly sorted,
regardless of their reported grain size.

po g as a measure of sorting. To check the validity of the charted sorting pattern
in the bay as determined by So, a chart (not figured) was prepared using the standard
deviation (¢o g) measures for the various stations. Inasmuch as ¢o g is based on
more of the cumulative curve, this value is considered a good criterion for evaluating
the So contoured chart, as well as for charting dispersion. As with the So (Trask
index) chart, the major fine sand areas are also outlined as regions of best sorting.
Within the area defined by the 0-50 ga g contour, a secondary region of exceptionally
well sorted fine sand was located parallel to the coast, and about 4 km. offshore
from Borth.

Even with an expanded coverage of the cumulative curve, no pronounced differ-
ences were noted in the distribution pattern of the poorly sorted deposits. The
rapid facies change from well sorted to poorly sorted sediments became more notice-
able, particularly in such areas as the inshore deposits from off Towyn northward
to off Ty-wen, and in the sarn-associated deposits bounded by the 2 ¢ Mz interval.
Estuarine deposits are found to be as well sorted using ¢o g as when using So values.
In all, no change is made in the basic sorting plan by the use of da g.

Intra-measure relationships. It is clear from Text-fig. 7, where ¢ Mz is plotted
against ¢o g, that with rare exceptions, the best sorted deposits are those in the
fine sand class. In fact, careful inspection shows that the best sorting is found in
the middle range of the ¢ 2 interval where ¢o g values as low as 0-15 and 0-20 are
reported. With increasingly larger grain size, even within the ¢ 2 interval, there
is a corresponding decrease in sorting.

The medium-sands, 1 to 2 ¢ Mz, have a mean ¢o g value of about one unit, while
the coarse and very coarse “‘ sands’ exhibit very poor sorting. There are few data
for granule and gravel sediments, but the largest size gravel samples show an indica-
tion of approaching reasonably fair sorting, with a mean of just over 1¢o0g. This
last observation suggests that for some of the gravel deposits, there is, at least
partial dynamic equilibrium within the physical energy environment. As can be
seen from the chart, the poorest sorting is found in the coarser sands, while the best
sorting is found in the finer sands. Paucity of data for ¢ Mz values below 3-0
precludes establishing significant trends for the very finest grained deposits.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 43

The most pronounced variations in skewness are found in the coarse and very
coarse sands. Since cumulative curves of the fine sands are almost symmetrical,
there is no preferred tailing-off of the curve in the defined limits for them. From
about the upper range of medium sands through the very coarse sands, Sk values
show a decided skewness with asymmetry tailing-off in the coarser fractions. How-
ever, for those deposits having ¢ Mz values near ¢1, the skewness is toward the
finer fractions.

Fic. 7. Correlation of grain size (f Mz) with sorting or dispersion (¢0 g.). Note that the
fine sands are the best sorted ones.

The correlation of ¢ Mz with ¢ M50, the mid-point datum, (Text-fig. 8) suggests
that while there is a good linear relationship for the fine and very fine sands and
silts, the coarser classes do exhibit scatter. Most of the coarser interval values
vary from the complementary ¢ Mz point by about one g-unit. Of the entire suite
of samples, only a few would be grossly misclassified if their textural nomenclature
were based solely on Md (6 M50). In practice, of course, most workers reporting
Md usually report corresponding So values, which would signal the reader to watch
for size dispersion and, thus, possible size term discrepancies. It is suggested,
therefore, that using caution regarding the coarse sand Md values, comparisons
can be made between these ¢ Mz classified deposits and others classified solely on
the basis of é M50 (Md) data. Similarly, Gripenberg (1934, p. 52) said, ““ The median
alone is a very good single characteristic of a sediment’. In short, since Cardigan

44 SEDIMENTATION IN NORTHERN CARDIGAN BAY

Bay fine sands are very well sorted, the median value may be used with confidence,
when making comparisons.

Summary discussion on texture. The deposits of this part of Cardigan Bay consist
of well sorted fine sands, and poorly sorted coarse (gravelly) sands, although some
few gravel deposits are well sorted. The finest sizes of the 2-3 ¢ Mz interval are
located close inshore, while the coarser grained sands are found farther seaward.
Texturally, there is no trend suggesting that fine sand is being delivered to the bay
through the estuaries. The fine sands and some gravel/cobble deposits are in equili-
brium with the prevailing environmental energy regime.

Fic. 8. Correlation of ¢ Mz with ¢M5o0. For the fine sands it makes little difference
whether ¢ Mz or ¢M50 values are reported even though the former statistic is based
on three points on the cumulative curve, and the latter on only one. While ¢ Mz is
obviously more desirable, the median diameter provides a good index of size for most
bay sediments.

The bay deposits may be described equally well using quartile measures, as by
using ¢@ based measures. For general comparisons, the size measures of ¢ Mz, ¢ MI
and ¢ M50 (Md) are interchangeable, thus permitting textural comparisons with
other similarly sorted marine deposits for which Md values alone are reported. The
writer suggests, however, that the more modern usage of such measures as ¢ Mz and
go g are to be preferred over quartile based terms in reporting textural descriptions
of Recent sediments.

Spencer (1963) in a critical review suggests that all deposits are made up of one,

SEDIMENTATION IN NORTHERN CARDIGAN BAY 45

or more, of three basic grain-size populations: gravel, sand and clay. Accordingly,
coarse sands and granule populations and silt populations are virtually non-existent
as single genetic types, and their presence is accounted for by resolving a bimodal
mixture into a single statistical entity, even if the curve does not exhibit apparent
bimodality. Further, under conditions of high energy, little or none of the clay will
be deposited, and the median, over a limited range, will serve as an indicator of the
physical energy of the depositional environment. While it is too soon to determine
how much of Spencer’s population mixing concept will prove acceptable, and more
data are needed to establish it on a firm empirical basis, to establish its validity is of
paramount importance to economic geologists in their exploration for sandstone
petroleum reservoirs. Likewise, sedimentologists, in their investigation and inter-
pretation of Recent clastics and clastic rocks, should find his concept particularly
useful for interpreting environmental energy systems. The present study appears
to confirm the mixing concept; indeed, the conclusions reached in the present study
have been tempered by our knowledge of Spencer’s findings.

In general, the sand facies in northern Cardigan Bay present no difficulty in chart-
ing, or in defining their textural nature. However, the coarse sands and the very
coarse sands, charted according to ¢ Mz measures, must be considered for what
they are, 1.e., mixtures of sand and either granules or gravel. Reference to Text-fig.
4, the chart of textural facies, shows very coarse sands and coarse sands extending
northward from the seaward end of Sarn Wallog. If we were to attempt an environ-
mental interpretation for this particular area (St. 190, IQI, 192, 201, 15 and 202)
based only on the representation provided by the chart, the suggested energy-
sediment relationship would not be particularly important, the reason being that
the bottom current energy needed to transport coarse and very coarse sands and
that needed for fine and medium sands is much the same (Hjulstrom, 1939). Indeed,
slightly less current (for transportation) is needed for the coarse sand according to
the Hjulstrom curve. However, in considering the same area, but in the light of
the Niggli scheme (Text-fig. 5), it is readily apparent that bimodal deposits of sand/
granules, and sand/gravels are present. If we accept Folk’s (1961) observation that
the presence of even a small amount of gravel is significant indication of increased
energy, then we must conclude that the mixed sand and gravel deposits extending
northward from the end of Sarn Wallog are transported northward by relatively
strong tidal currents. Moreover, considering the large amount of gravel present
(29°4% at Station 190, for example), it is suggested that tidal currents are three
times, or more, stronger in the mixed gravel and sand zone than outside it. This
linear extension of coarse deposits appears to be related to a dispersal pattern whereby
coarser deposits are removed from the end of the sarn, or transported around the
sarn, and then carried northward and mixed with sand.

In the middle part of the present survey and west of Sarn Bwch, a similar situation
exists. There, very coarse sand, according to the ¢ Mz chart (Text-fig. 3), is more
realistically termed sandy gravel, and sandy granules in keeping with the Niggli
scheme (Text-fig. 5). A similar relationship is also reported for very coarse sands
and granules, and the sandy granules and sandy gravel in the region north of Sarn
Bwch in mid-bay.

40 SEDIMENTATION IN NORTHERN CARDIGAN BAY

These examples are few in number and bimodal deposits constitute a small areal
portion of the bay floor ; nevertheless, they are critical indicators of the high energy
bottom current regime. The present data are such that the sarn source interpreta-
tion is not without an alternative. Careful study of the reported textural data and
the charts suggests that the granule and larger size fragments are most probably
derived from the erosion of the outer ends of the sarns, but it is possible that a part
of the granule/gravel fraction of the mid-bay bimodal sediments may be material
transported around the sarns from an ultimate source somewhere along the southern
bay coast.

Vo OPTICAL PET ROGRE Pay

(a) Petrographic classification of grain types

After studying the thin sections of bay deposits as a whole, and ascertaining the
common grain-types an empirical system of nomenclature was devised. This was
done mainly because quartz studies (Krynine, 1940; Tuttle, 1952; Folk, 1961 ;
Blatt and Christie, 1963) suggest that a given quartz type may originate from more
than one kind of rock, for instance quartz with undulating extinction may be igneous,
metamorphic, or reworked sedimentary in origin. For quartz the writer has chosen
a binomial system adapted from Folk’s (1961) scheme, and this has been combined
with other minerals, in a classificatory scheme without genetic implications. Such
an approach is well suited to the petrographic study of Recent sands because it
allows dispersal charts to be drawn without needing a prior provenance study to
establish those types which might be expected offshore, and it provides a flexible
framework in which new types, particularly those with inclusions may be added.
Detailed descriptions of the various grain categories, and examples of their use in
provenance interpretations are to be found in Folk (1961) and Todd & Folk (1957)

Quartz was assigned to one of fifteen types in addition to secondary quartz over-
growths; the latter, a special type, is set off from the others. A binomial designa-
tion was used in which the first term was based on the characteristic extinction
of the grain, and the second term was based on the presence or absence of inclusions,
and the kind of inclusions when present.

In the present study, six extinction types are used in quartz classification :

1. Straight extinction. It is defined as simultaneous extinction of all parts of the grain, ie.,
the grain is “‘ darkened ”’ uniformly (cf. BM 1964, 173 & 174).

2. Slightly undulose extinction. A definite, but broad, extinction shadow is observed to move
across the grain, usually requiring up to five, or slightly more, degrees rotation of the stage
(Folk, 1961, p. 71), cf. BM 1964, 99 & 287. N.B. The grains are randomly orientated.

3. Strongly undulose extinction. This type of extinction usually requires somewhat more than
five degrees (up to 20° in some cases) of stage rotation for the extinction shadow to move
from one extremity of the grain to the other. In practice, one easily recognizes strongly
undulose extinction by the pronounced, narrow linear shadow which sweeps across the grain.
In this extinction class, as in the previous two, reference is made to a single sand grain in
crystallographic continuity (cf. BM 1964, 130 & 286).

4. Semi-composite extinction. In this category, as well as in the two which follow, reference is
made to detrital fragments of quartz composed of two or more crystals, i.e., polycrystalline.
Here the fragment is composed of joined, contact-to-contact sub-grains, which possess very

SEDIMENTATION IN NORTHERN CARDIGAN BAY 47

close optical orientation. As the stage is rotated, each grain becomes extinct in progressive
order across the fragment, with extinction being straight, or only very slightly undulose for
each grain. The sub-grain contacts within the fragment are not crenulated, and the extinc-
tion shadow does not cross grain boundaries within the fragment (cf. BM 1964, 100 & 102).

5. Composite extinction. This category includes fragments composed of two or more sub-grains,
or crystalline individuals, with grossly different, usually irregular extinction. There is no
preferred pattern ; extinction occurs first in one grain, then in another, randomly across
the fragment. Borders of the grains within the fragment are not crenulated, and the extinc-
tion of each individual is straight, or only very slightly undulose (cf. BM 1964, 172 & 282).

6. Composite metamorphic extinction. This classification type includes polycrystalline fragments
which exhibit strongly undulose extinction. In most cases the extinction shadow is pro-
nounced as it sweeps across the fragment. Most fragments possess sub-grains with crenulated
contacts, but this is not an absolutely necessary criterion for classification.

The second criterion for empirically classifying quartz grains is the nature of the
included material, i.e., vacuoles, small crystals of other minerals, etc. While, in
theory, the number of variables of this sort appears to be large, in practice only a
few inclusion types are found in any one region. Such was found to be the case for
the sands in the northern part of Cardigan Bay, where only four inclusion types
occurred in sufficient numbers to provide statistically reliable counts. Preliminary
study of the thin-sections revealed inclusions of vacuoles, microlites, rutile needles
and chlorite flakes. It was obvious, too, that many of the grains, in fact the majority
at some stations, did not possess inclusions and, thus, were categorized as “‘ clear ”’
(cf. BM 1964, 95 & 98). Common inclusions found in the Cardigan Bay sands are
as follows :

1. Vacuoles. These are small liquid filled chambers, usually globular in shape and which
appear to be brownish in colour. In most instances, the vacuoles are spatially distributed in
a plane, or several planes, throughout the grain. Under medium power magnification, the
vacuoles appear as linear traces, but subsequent inspection under high power magnification
reveals their planar distribution. Some quartz grains possess vacuoles which are scattered
randomly through the grain. Such types are moderately common (cf. BM 1964, 236 & 252).

2. Microlites. These are, by present definition, all mineral grains within the quartz fragment,
except chlorite and rutile needles. Some variation was observed in the types of included
minerals, but tourmaline, opaque unknowns and zircon, as ultra-small, euhedral crystals
were commonly noted. Other minerals were rarely present (cf. BM 1964, 89 & 134).

3. Fine (rutile) needles. Inclusions of very fine, hair-like needles of rutile are found in some
quartz grains. The extremely small size of these crystalline inclusions is a useful clue
to their identity, and in this study rutile needles and the descriptive term “‘ fine needles ”’
are used interchangeably (cf. BM 1964, 158 & 276).

4. Chloritic inclusions. Small flakes of chlorite were observed in a few grains, but such inclusions
are not common. Care must be taken in establishing that the chlorite is actually within the
fragment (cf. BM 1964, 104 & 197).

Although detrital quartz formed the bulk of most sand samples other grain types
are considered important constituents of the inclusive classification, these being as
follows :

1. Secondary quartz. Several different types of grains, usually quartz, very rarely feldspar,

were observed with secondary overgrowths of quartz, in optical continuity and euhedral
in form (cf. BM 1964, 270 & 277).

48 SEDIMENTATION IN NORTHERN CARDIGAN BAY

2. Carbonate fragments. All calcareous grains have been assigned the collective designation of
“carbonate fragments ’’. Only rarely were lithic carbonate rock fragments identified, and
these were largely restricted to pebble deposits. The importance of this category is that it
provides data for estimating, albeit roughly, the amount of carbonate material added to the
sediments by biological processes in this particular marine environment (cf. BM 1964, 248
& 259).

3. Avenites. This category includes all rock fragments derived from sandstones and siltstones
(cf. BM 1964, 152 & 222).

4. Slate. All shale, slate and phyllite fragments are grouped under this single petrographic
classification. The small grain size of the majority of the bay deposits makes such assignment
a necessity, for while a small grain can be determined as a lithic fragment composed chiefly
of phyllosilicate minerals, little else about the original rock can be established (cf. BM 1964,
107 & 222).

5. Crystallines. Although certain of the crystalline rock types were readily identifiable in
granules and in large sand grains, absolute type identification within the fine sand sizes
was found to be limited to the occasionally encountered granophyre fragment. It was
found, however, that even in small grains the general crystalline rock nature could be deter-
mined with reasonable assurance even though a specific clan name could not be assigned.
Under such circumstances, all crystalline rock fragments were assigned to this common
category (cf. BM 1964, 198 & 307).

6. Feldspar. Inasmuch as feldspathic grains most commonly occurred in sizes smaller than the
host, or quartz, fraction, thus creating some difficulty in both statistical counting and type
identification, it was decided to assign all feldspathic grains to one common group. X-ray
diffraction data, also reported, are considered to provide a more accurate determination and
measurement of the feldspars; furthermore, classification schemes, such as Pettijohn’s
(1957), require only that total feldspar be pole assigned (cf. BM 1964, 180 & 273).

Counts of the several grain types were normalized, and relative percentage values
were assigned to each classification category. Although counts were originally
recorded for the various quartz extinction types with microlites, fine needles and
chlorite flakes as inclusions, the totals for each of these three categories were sub-
sequently grouped. No attempt was made to determine accessory minerals or rare
grain types.

(b) Dispersal of grain types and vectors

Using petrographic data obtained from grain counts, dispersal charts for the
various grain types were drawn and studied. Such graphically displayed information
not only shows the quantitative distribution of each individual type, but it also
provides clues to the source and to the dispersal direction. Pettijohn (1957, pp.
525-528) lucidly summarized the concept of dispersal by saying: “ Although the
products of weathering may be widely dispersed, there is a tendency for those pro-
duced in any one area to move out in a systematic manner and to be spread out over
a rather well-defined region or area. Such a region or area, whether large or small,
is a dispersal shadow deposited downcurrent from the generating or source area’.

The dispersal vectors presented are proposed as reliable indicators of transport.
The direction of each vector was determined by drawing a line along the axis, or
axes, of major concentration for every contoured salient, with the vector origin in
the region of maximum concentration. The length of the vector is proportionate
to the distance each salient extends geographically, and it normally represents the

SEDIMENTATION IN NORTHERN CARDIGAN BAY 49

distance from a local source to the zone of diminution by mixing. In some cases,
the vector extends slightly beyond the limiting contour of the salient, and this
extension is based upon data for samples beyond the contour. In the instances of
certain charted trends which suggest zones of transport, but for which no definite
directions of diminution are apparent, “ headless”’ lines are drawn. Obviously,
critical judgment is an important part of the vector approach to dispersal study.
In the following pages an attempt has been made to point out for each petrographic
type those salient trends, or dispersal patterns which are important and most likely
to lead to the recognition of sediment sources and zones of effective transport.

(x) Quartz, straight extinction, clear. Study of its charted distribution (Text-fig. 9)
suggests that there are five areas containing anomalously large amounts of this
grain type. First, the area along the western margin of the survey, delimited by
the 30% contour, which extends from Station 250 (32%) in the north to Station
260 (47%) in the south. The stations within this area are largely in zones of medium
grained sands, and high values at Stations 260 and 269, both above 40%, suggest a
source for these quartz grains outside and to the south of the study area. Second,
an area west of Towyn defined by Stations 58 and 276, both over 40%, and nearby
stations with over 30% clear, straight extinction quartz. Another small area near
Towyn, also with values over 30%, is in the vicinity of Stations 30 and 277 and is due
south of Towyn. Third, two areas which in all likelihood are contiguous zones of
sedimentation, are the area south of Sarn Wallog (Stations 46, 47, 263, 264, 265 and
267), and the area just west of the sarn bounded by Stations 48, 261 and 262. The
dispersal pattern of clear, straight extinction quartz in this general region, like that
farther west, suggests transport from the south part of Cardigan Bay. Fourth, a
combined littoral-estuarine dispersal zone extending from near Station 132, west of
Ty-wen, thence northward along the coast, around the spit at Barmouth and up the
south side of the Mawddach estuary to the vicinity of Station 140. All the stations
within this area, except 41 (37%), have values between 30 and 33%. Fifth, the
large U-shaped area which is developed in the extreme northern end of the survey,
i.e., in the corner formed by Sarn Badrig and the coast with a mid-bay, south-
westerly extension through Stations ror to 274. This area, both in its shape and
in its higher content of clear, straight extinction quartz, suggests that this quartz
type is being supplied from beyond the sarn, and in a considerable quantity.
While no petrographic data are yet available for sediments north of Sarn Badrig,
it appears that a source there is likely, and some detritus is transported over the
sarn at shoaling depths. There is no field evidence to suggest that the Afon Ysgethin
provides detritus in considerable volume. It is unlikely, then, that this particular
grain type could be introduced by such a small stream in an amount sufficient to
account for the anomalous concentration in this northern area.

The irregular, elbow-shaped pattern of dispersal is explained as the composite
effect of the two sets of tidal currents, each varying in magnitude and direction with
the respective tidal period.

(2) Quartz, slightly undulose extinction, clear. Although Folk (1961) and Blatt &
Christie (1963) have offered opposing arguments on the origin of quartz with undulose

50

Fic. 9.

SEDIMENTATION IN NORTHERN CARDIGAN BAY

(oy Shs el Gt) So a
|
KILOMETRES

e ‘ Barmouth 20

O

Distribution of clear quartz grains with straight extinction.

z Sef.
See
Y¥140

NV. es
. cg) cee ae
3

BA — BADRIG
BW — BWCH
WA — WALLOG

Contour interval 10%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY

Orde s2ees 4) S678 10_11
—— a ——
KILOMETRES

i: © Ty-wen

e
SARNS
aN {7/1 Peet's BA - BADRIG
eae cares ert e fia: BW- BWCH
eS 40262 To -% : WA- WALLOG

= axe
NoNercee 0° : asl 4
ue) es ain

51

=)

S|

Fic. 10. Distribution of clear quartz grains with strongly undulose extinction.
interval 2:0%.
MINER. 2, 2.

Contour

52 SEDIMENTATION IN NORTHERN CARDIGAN BAY

extinction, it is not necessary to establish conclusive evidence supporting either
viewpoint in order to characterize such grain types empirically.

This grain type is concentrated in the shallow, inshore sands and its dispersal
from the eroding cliffs south of Borth is confirmed by the data for stations in that
area. In general, this quartz variety is transported northward and it is unrelated
to either river system. The slight enrichment in deposits along the south side of
the Dyfi estuary is the result of inward transport from the sea and subsequent
deposition in the marsh flat. A small amount of this quartz type may be introduced
from beyond Sarn Badrig, probably transported through the trough on the landward
end of the sarn.

(3) Quartz, strongly undulose, clear. In common with slightly undulose, clear
quartz, the strongly undulose variety shows a definite dispersal development in
fine sands inshore between Borth and Sarn Bwch (Text-fig. 10). The 2% contour
encloses several smaller areas wherein values exceed 4%, and are as high as 8% at
one station [53]. Although sample station control does not allow one to bring all
4% contours together, it appears reasonably certain that these smaller, higher per-
centage areas are contiguous. Another area with high values, which appears to be
a zone of transport, is located west of the outer end of Sarn Wallog. Stations 48
(8:1%), 262 (8:1%), 49 (5°4%) and 260 (5:3%) establish the charted concentration,
and when combined with Stations 263 to 267 inclusive, suggest an origin south of
the survey.

A typical dispersal shadow is developed in the northernmost part of the survey
based on data for Stations 238 (7:-3%) and 239 (10:0%). In referring to Text-fig.
ro, the progressive diminution of strongly undulose, clear quartz in a southerly
direction is suggestive of a source north of Sarn Badrig. Inasmuch as Stations 90

(5:0%) and 243 (2:-4%) are only 2 km. apart, it is reasonable to assume that the
dispersal extension of this quartz type southward to Station 38 (4:9%) is continuous
and related. Presumably, the undulose grains dispersed northwesterly from off
Aberdyfi, and southeasterly from the northern source (through Stations 245, IOI
and gg) provide the coalescing pattern which then extends northwesterly toward
Stations 253, 252 and 250.

There is no evidence to suggest that strongly undulose, clear quartz is dispersed
from within the Mawddach estuary system. Three selected stations at the mouth
of the estuary in the vicinity of Barmouth, numbers 135, 143 and 123, while textur-
ally similar to the other nearby sands, do not exceed 0-9%. Three stations within
the upper estuary with values exceeding 2°{—Stations 125, 140 and 127—do not
negate the preceding observation, in that all three were sampled in marsh sands,
sediments which were deposited during periods of highest tides, as determined from
field evidence. Several samples from the Dyfi estuary contained strongly undulose,
clear quartz in amounts between 2 and 3°5% ; such is thought to have been brought
in from Cardigan Bay. None of the river samples in this region contained strongly
undulose, clear quartz.

(4) Quartz, semicomposite extinction, clear. One of the most informative charts
to be prepared during the course of this study is that drawn on the basis of data

SEDIMENTATION IN NORTHERN CARDIGAN BAY 53

for semicomposite extinction, clear quartz (Text-fig. 11). The major dispersal
pattern for this particular grain type is such that no basis exists for an alternative
interpretation. Reference to the chart shows that semicomposite, clear quartz,
in amounts over 1%, is effectively restricted to an area extending northwesterly
from the cliffed coast near Borth, and terminating near the seaward end of Sarn
Bwch. Its distribution, as delimited by the contour lines drawn for values between
I and 6%, provides an almost ideal example of the dispersal shadow as defined by
Pettijohn (1957). A minor local source of this quartz type is the coastal area of
eroding cliffs between Borth and Sarn Wallog. No semi-composite, clear quartz
was found in stream bed deposits of the Dyfi or the Mawddach. In summary,
semicomposite, clear quartz is restricted chiefly to that part of the study area south
of Sarn Bwch, and the main source appears to be, or be near, the eroding sections
between Aberystwyth and Sarn Wallog.

(5) Quartz, composite, clear. Unlike the preceding variety, composite, clear quartz
is primarily dispersed in the northern part of the study area (Text-fig. 12). There
is, also, a partial development along the western margin of the area between Stations
259 and 118, suggesting dispersal from beyond the westward limit of this study.
The contour which delimits this local increase is based on four stations with values
over 2%, and one station with a value of 4:5% (257).

Dispersal in the far northern part of the area is confined to three almost parallel
extensions, which have a common zone along the coast between Barmouth and
Mochras. They may not be related, however, to a common source and, thus, are
considered separately. The first and most pronounced dispersal trend is that aligned
from Station 79 near the end of Sarn Bwch northward through Stations 98, ror,
245, 219 and 236. This elongate zone contains two stations, 98 and ror, with values
between 4 and 6%. It is suggested that greater amounts in the northern part of
this dispersal lineament have been masked or diluted by the influx of other varieties
being brought in from north of Sarn Badrig. The second dispersal zone extends
from the cliffed coast west of Ty-wen northward and parallel to the coast to the
Mawddach estuary mouth. Littoral sands along the coast in this area are somewhat
higher (by about 2%) than the deeper water deposits seaward from them. This
observation together with the recognized northward drift of inshore sands is suggest-
ive of a source in the vicinity of Stations 132 and 133. Glacial deposits are being
eroded nearby, and they probably serve as the local source for composite, clear
quartz found in these nearshore and littoral sands. The third dispersal trend is that
based on three stations, 272, 222, and 221, parallel and immediately adjacent to
Sarn Badrig. These stations appear to lie in the zone of fine sand transport
located close to the sarn. The dispersal patterns for this zone and the mid-bay
zone appear to be related to a common source of composite, clear quartz some-
where north of Sarn Badrig and beyond Mochras. Certain samples within the
Mawddach estuary (142, 141, and 127) with amounts between 2 and 3-7 % do not, in
themselves, suggest that the Mawddach river system is providing composite, clear
quartz.

Briefly, composite, clear quartz is primarily dispersed in the northern portion
of the study area, and it appears to be derived from the glacial deposits near Ty-wen

54 SEDIMENTATION IN NORTHERN CARDIGAN BAY

AS 160077 18h Sei) il
Ft — = ——|
KILOMETRES

‘Bar mouth 1%:

BA — BADRIG
BW — BWCH
WA — WALLOG

Fic. 11. Distribution of clear quartz grains with semi-composite extinction. Contour
interval 1:0%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 55

OLMi2a Gi 4 56 7 SS EMOlntt
[SS SS SS a SS
KILOMETRES

SARNS
BA — BADRIG
BW- BWCH
WA - WALLOG

Fic. 12. Distribution of clear quartz grains with composite extinction. Contour interval
2055.

56 SEDIMENTATION IN NORTHERN CARDIGAN BAY

and an unknown source beyond Sarn Badrig. Although small quantities were found
in Mawddach estuary sands, their presence is not considered important.

(6) Quartz, composite metamorphic, clear. The dispersal pattern of this quartz
type (Text-fig. 13) presents a readily understood distribution. On the basis of the
reported data, it is dispersed from the vicinity of the cliffs south of Borth, and is
distributed in a general northwesterly direction. It is unlikely that Stations 31
and 32 with between 2-0 and 3-2°% composite metamorphic, clear quartz constitute
a separate dispersal zone, for while sample coverage is not dense enough to permit
detailed contouring, this area is most likely connected with that to the south. Iso-
lated stations such as those found in the northern part of the survey (85, 92, 123,
etc.) do not appear to be part of any plan. In all, composite metamorphic, clear
quartz is being dispersed only in the offshore sands west of Borth, and probably
owes its origin to erosion of the coastal cliffs of Aberystwyth grits.

(7) Quartz, straight extinction, vacuole inclusions. ‘This is one of the most common
quartz types present in the Cardigan Bay sands, and it is difficult to establish definite
dispersal trends and even more difficult to conclude where it may originate. Of
course, the fact that any specific grain type is ubiquitous may well suggest it is
part of the reworked detritus, distributed earlier by the transgressive sea. In
some respects, this grain type may be thought of as one which forms a part of the
detrital ‘‘ background ’’, or pre-existing clastic matrix. Nevertheless, there is some
suggestion based on data for stations within the Mawddach estuary, that a source
upstream for this quartz type, at least in very small quantities, is possible. The
data are not, however, conclusive.

(8) Quartz, slightly undulose, vacuole inclusions. This grain type occurs in amounts
about twice that for the previously reported clear variety. The vacuole type,
however, presents a more meaningful distribution in that its dispersal is within
an elongate, contiguous zone paralleling, but offshore from, the coast. Inasmuch
as slightly undulose, vacuole inclusion quartz is relatively common in these near-
shore sands, it was charted for quantities above 10% in order to delineate its dis-
persal (not figured) in the northern part of Cardigan Bay. One source appears
to be the eroding cliff section just north of Sarn Wallog. Several stations in this
area, notably 150 (15:5%) 151 (16:6%) and 17A (12-7%), provide substantial evi-
dence for concluding that a dispersal centre exists there. Another source of this
grain type is the eroding glacial debris in the vicinity of Towyn, for immediately
offshore Stations 66 (17:0%), 67 (13:7%), 65 (15:1%) and 32 (124%) all possess
greater amounts of this grain variety than do stations surrounding them. In fact,
these stations provide a sound basis for contouring the lobed distribution outward
and southward from the Towyn area. In summary, a major dispersal originates in
the cliffed coastal area south of Borth, with a subsidiary source in the vicinity of
Towyn. There is no dispersal area of significance in the northern part of the survey.

(9) Quartz, strongly undulose, vacuole inclusions. With a distribution pattern
somewhat resembling the previous variety, the strongly undulose, vacuole inclusion
quartz is distributed in a linear irregularly shaped band extending from Borth
northward along the coast to the inshore end of Sarn Badrig. Inspection of the

SEDIMENTATION IN NORTHERN CARDIGAN BAY 57

(tes a SG. U7 Bee iO at
KILOMETRES

\: Be

DQ Barmouth of

aaa Ve
ij

BA — BADRIG
BW - BWCH
WA — WALLOG

Fic. 13. Distribution of clear quartz grains with composite metamorphic extinction.
Contour interval 2:0%.

58 SEDIMENTATION IN NORTHERN CARDIGAN BAY

dispersal plan (Text-fig. 14) shows that significant amounts are also found in the
far western part of the present survey. Inasmuch as the highest values reported
are for those stations west of the Borth beach, Stations 21, 54, 51 and 60, one may
conclude that the southern part of the survey is a centre for local dispersal of this
grain type. Here again it seems evident that the cliffed coast near Borth serves as a
centre from which much of this quartz variety is dispersed in northern Cardigan Bay.
However, southward directed tidal currents entering the area from Tremadoc Bay,
over the landward end of Sarn Badrig, may transport additional amounts of this
grain-type into the area. This is based upon data for Stations 238 (8-8%), 237
(7-4%), 239 (7°4%), 24X (7°6%), 236 (68%) and 244 (8-2%). As the geographic
arrangement of these stations is such to provide a dispersal shadow with the data
maxima farther north, it is concluded that strongly undulose, vacuole quartz is
introduced from Tremadoc Bay. Neither the Afon Mawddach nor the Afon Dyfi
transport this type of quartz to Cardigan Bay. Likewise, none of the smaller streams
are active carriers of this quartz, nor is it found in quantities over 5% in the littoral,
or beach sands with the exception of Stations 173 and 133. In regard to this obser-
vation, i.e., the lack of appreciable amounts of undulose quartz in the littoral de-
posits, it is indeed possible that some strained varieties of quartz are susceptible to
erosional fracturing under high energy conditions prevailing in the surface zone, and
even offshore where large waves dissipate their energy against the bottom.

(10) Quartz, semi-composite, vacuole inclusions. Relatively large amounts of
semi-composite, vacuole quartz reported for Stations 9 (9:2%), 17 (8:1%), 18 (7:7%),
6 (8-8%) and 180 (7-4%) suggest that this variety is derived from a source in the
coastal area just south of Borth (Text-fig. 15). Indeed, even those littoral samples
largely composed of lithic grains (144, 148 and 149) contain a higher amount of this
quartz than might otherwise be expected, due, of course, to their proximity to the
local source. If one follows the general dispersal trend enclosed by the 4%
contour, a noticeable correlation of distribution with the northward axis of the tidal
current system is seen. Thus, by beginning near the cliffs and tracing the 4%
contour one may draw a reasonable path for transport northward to its intercept by
Sarn Bwch. At Sarn Bwch the dispersal turns westward and continues to the end
of the sarn. As outlined in previous discussions on tidal currents and textural
distribution, the present dispersal plan likewise follows a similar pattern by extending
a dispersal halo north of Sarn Bwch, and terminating (based on the 2% contour)
in the vicinity of Station 88. There is, of course, a westward dispersal of this quartz
type which is somewhat less sharply defined but still obvious from the evidence of
Stations 22, 76, 116 and 2or, all with values above 4%.

Fine sand is being transported into the estuaries from the bay. The present distri-
bution chart (Text-fig. 15) supports this significant observation. Reference to the
chart shows that a delimiting 294 contour irregularly crosses the Dyfi estuary,
seaward of which values in excess of 6% are contoured. On the other hand, there
is, east of this contour, a noticeable decrease for the several stations from 1:8% at
Station 169 to nil further to the east

An apparent, but minor, dispersal suggesting an origin beyond Sarn Badrig and

Fic.

1

14.

SEDIMENTATION IN NORTHERN CARDIGAN BAY

An ao G/B OEnOeN
KILOMETRES

59

\ oes il A *

a)

BA — BADRIG
BW —- BWCH
WA—- WALLOG

Distribution of quartz grains with strongly undulose extinction and vacuole
inclusions. Contour interval 5:0%.

60 SEDIMENTATION IN NORTHERN CARDIGAN BAY

showing transport over the far outer edge of the sarn, is developed on an axis through
Stations 216 (8-9%), 231 (5:7%) and 224 (3:5%). Were it not for the progressive
increase of this quartz variety toward the end of the sarn, one might offer the alter-
native of connecting Stations 87 and 224 in an area of coarse clastic deposits which,
texturally at least, appear to be related.

(11) Quartz, composite, vacuole inclusions. Although this quartz variety appears
to be dispersed away from the glacial exposures near Ty-wen in minor amounts,
its main zone of enrichment is along the shore in beach deposits. Some local ac-
cumulation occurs on the north side of the Dyfi estuary, but seems unimportant.
The concentration of this quartz variety in the beach and nearshore zone suggests
that it is primarily transported by drift currents. Indeed, it may be simply a
transient grain variety with a source elsewhere in the bay. Its distribution offshore
cannot be rigidly defined or interpreted.

(12) Quartz, composite metamorphic, vacuole inclusions. For the majority of the
samples from northern Cardigan Bay, composite metamorphic, vacuole inclusions
quartz was found to be present in amounts less than 2%; in about half of them
this variety accounted for less than 1% of the grains. In most of the offshore
area it is randomly distributed, and the data do not suggest a definite dispersal
pattern except in the western margin of the survey. This exception, a lobe-shaped
zone in the west-central portion of the survey, is recognized on the basis of data for
Stations 118 (2:2%), 195 (2°7%), 196 (5°1%), 247 (3°0%), 271 (2°7%), 272 (42%),
229 (2°4%) and 224 (4:1%). Inasmuch as these stations are near the margin of
our present survey and as no data are available for the deposits farther seaward,
the present dispersal plan suggesting a source, or zone of transport, beyond the
seaward end of Sarn Badrig must be considered a provisional one.

Two local areas of concentration are the littoral zone paralleling the cliffs south
of Borth and the littoral zone between Aberdyfi and Sarn Bwch. There is no
evidence to suggest that this quartz type is transported to the sea through either
the Dyfi estuary or the Mawddach estuary.

(13) Quartz, all extinction types, microlite inclusions. Because quartz grains con-
taining microlites were largely restricted to the straight extinction class, the author
has grouped them in a single category for this study. A dispersal plan which is
both easily determined and straightforwardly interpreted is evident (Text-fig. 16).
Furthermore, the contoured plan is developed with the characteristic dispersal
shadow form (Pettijohn, 1957) with those stations having the largest percentages
being restricted to a narrow zone northwest of the cliffs near Borth. The apparent
axis of dispersal is northwest from the cliffs to the end of Sarn Bwch where a second-
ary zone extends westward. From the end of Sarn Bwch, or near that general
vicinity, a dispersal trend extends northeastward to within the vicinity of Station
217. Some microlite-bearing quartz is dispersed northward from near Towyn.
For the latter dispersal, which joins the main northward transport zone, a source
near Towyn, probably the glacial debris exposed there, is suggested. This type of
quartz is not being delivered to the bay through the Mawddach and Dyfi estuaries,
nor is it being introduced from beyond Mochras point in the north.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 61

OMin2te si 425) 06.272 18eoreOei
— = — eas
KILOMETRES

BA — BADRIG

wA> ca i. BW — BWCH

WA — WALLOG

2 e So.
AL , Bianervetwyen
ae ip erystwy

Fic. 15. Distribution of quartz grains with semi-composite extinction and vacuole
inclusions. Contour interval 2:0%.

62 SEDIMENTATION IN NORTHERN CARDIGAN BAY

(14) Quartz, all extinction types, rutile inclusions. For most of the estuary and bay
deposits, quartz grains with rutile inclusions (“ fine needles ’’) occur in small amounts
(Text-fig. 17). Indeed, only eleven stations have values exceeding 2% and for
only two stations do values exceed 3%. Within the southern half of the survey, a
dispersal of this quartz variety extends irregularly northward from the outer area
of Sarn Wallog. While this might be the result of submarine erosion of westward
extending sarn deposits and their subsequent transport northward, a more likely
explanation may be a distant source south of Sarn Wallog. The few data from
this southern area do not permit a firm conclusion. In the northern part of the
survey adjacent to the outer Sarn Badrig region, a second centre for dispersal of
rutile-bearing quartz is developed. There are no data which suggest that this
quartz variety is being delivered through either of the two major estuaries. In
short, its chief source area is apparently beyond the limits of this survey.

(15) Quartz, chloritic inclusions, all extinction types. Chlorite-bearing quartz is
dispersed from two source sites. The first, and more important, are the exposures of
glacial debris along the coast south of Barmouth. The second is indicated by a
weakly developed pattern along the Borth beach. It is probable that the chloritic
quartz from these two areas is of different origin. Some patchy development of
this quartz variety occurs offshore, but no obvious pattern can be discerned.

(16) Quartz, secondary overgrowth. In spite of the paucity of grains with over-
growths (0-0 to 2°7%, av. 0°5%), it is possible to establish two dispersal plans (not
figured) for the bay sands. The first is an elongate zone extending northward from
Sarn Wallog in mid-bay and its northward extension beyond Sarn Bwch, and the
second is the littoral area between Sarn Bwch and Eglwys Llanaber, also including
the lower Mawddach estuary. In the latter instance, it appears that grains with
secondary quartz overgrowths are transported into the estuary from littoral deposits
between Ty-wen and Barmouth. With the exception of littoral sands north of
Sarn Bwch which may come from the erosion of glacial debris along the coast there
(Pl. 2, fig. 2), the major dispersal of this grain class appears to originate in the
southern part of the study area, perhaps associated with Sarn Wallog, but more
likely from a distant source beyond the southern limit of this survey.

Arenite fragments. The distribution of arenite fragments (Text-fig. 18) is based
solely on the sampled and analysed sand deposits, and on certain coarser clastics
which are not part of the sarn accumulations. The sarns are composed of gravels
and boulders. This discussion relates to the relatively more mobile sediment cover
of the Cardigan Bay bottom. The same qualification applies to comments on certain
other distributions, namely slate fragments and crystalline rock fragments.

The most noticeable pattern and, perhaps, the most significant one, is that which
extends northward from the cliffs at Borth and is eventually directed seaward in the
vicinity of the mouth of the Dyfi estuary. This dispersal pattern suggests an origin
for much of the arenite debris in the vicinity of the cliffs just south of Borth, an
observation which, incidentally, is supported by field study (PI. 2, fig. 1). Stations
along the littoral zone at the base of the cliffs between Aberystwyth and Borth possess
not less than 21:5 % arenite fragments (Station 267), and at one station (180) their

SEDIMENTATION IN NORTHERN CARDIGAN BAY 63

content reaches 46%. Those stations at the base of the cliffs between Sarn Wallog
and Borth owe their enrichment of rock fragments to erosion of the immediate
cliffs (Pl. 2, fig. 1). A progressive diminution of arenite fragments is noted when
one considers the data for Stations 25 (141%), 17 (9°4%), 18 (86%), 205 (62%
and rr (4:3%). These stations are all in line seaward of Borth. A consistent
amount (between 10 and 11%) is reported for those samples along the beach front
between Borth (Station 150) and the outer spit (Station 153). The distribution of
arenite fragments in the lower Dyfi estuary and in the vicinity of its opening to the
sea at Aberdyfi suggests their dispersal into the estuary from the bay. Station 20,
some II km. west of Borth, with 37-2°% arenite rock fragments, is considered an
anomalous station, and its higher content is attributed to textural influence (¢ Mz,
—1-2) ; the same is true for Station 4 (6 Mz, —1-0). In most of the sands from
both the southern and northern parts of this survey, the arenite fragments are present
in amounts varying between 4 and 9%.

A second source is the glacial debris in the vicinity of Towyn, which shows its
influence westward for some 34 km. Station 171, immediately in front of the
eroding shore at Towyn, contains 24:5 °% arenite fragments, and Station 65, about
14 miles west of it, contains 13-3°%; thus, the lithic content diminishes rather
rapidly with increasing distance seaward. The control of this diminution may be
either the limited amount of material available, or it may be that lithic fragments
are broken into constituent sand grains as they pass through the high energy surf
zone.

In the northern part higher values for Stations 272, 216, 222, 221 and 215, all
between Io and 15%, suggest that the arenite grains in this area are derived from
Sarn Badrig. Likewise, random Stations 224, 226, 196, 182 and 183, with values
over 10 %, may be related to sarn erosion or to textural control, and need not neces-
sarily be part of a dispersal scheme. It should be noted that the significantly higher
arenite content of the several stream samples correlates with the larger grain sizes.
There is no evidence suggesting that arenite grains are being spread over the estuaries
by their associated streams.

Slate fragments. In the southern half of this survey two dispersal patterns are
of major importance in understanding the distribution of these grains, as well as
several of the chemical elements. The first of these patterns is the dispersal zone
from the cliffed area south of Borth. This area is delimited by 5 % values, although
within it sands with slate fragments in amounts less than 5% are known. The
stations in the latter area (51, 52, 4, 68, 1 and 2) show a wide spread of values below
5%. Station r with 0-8% and Station 2 with 2-4% suggest local masking by other
grain types, and Station 4 with no slate fragments may be an expression of textural
control. Station 20, approximately 10 km. west of Borth, is an anomalous station
within the area, since it contains 16:2 % slate fragments, and may represent an exten-
sion of the charted higher fragment content deposits westward some 5 km. In
general, however, the littoral deposits between Aberystwyth and Borth with values
exceeding 20% in places suggest the nearby cliffs as the immediate source.

Farther offshore, a pronounced tongue-shaped dispersal pattern extends north-

64 SEDIMENTATION IN NORTHERN CARDIGAN BAY

Ne
ee ese ee Gy ©) Osh LA (A
_—S— | x 2x
KILOMETRES Z S See
Za : ees
Ps 5 e N=
a “ye ° Wee ‘ 5 ;
a. iy o NE: ee
a & Svs i) eae eae
. *\ Barmouth : a

BA — BADRIG
BW — BWCH
WA- WALLOG

Fic. 16. Distribution of quartz grains, all extinction types, with microlite inclusions.
Contour interval 2:0%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 65

a i: Wa
—— = oe

9 10 11

KILOMETRES

SARNS
BA - BADRIG
BW - BWCH
WA -— WALLOG

Fic.17. Distribution of quartz grains, all extinction types, with rutile inclusions. Contour
interval 1:0%.

65 SEDIMENTATION IN NORTHERN CARDIGAN BAY

‘
O12 3 14.5 16.7 38 89 0m a
(Ss ee LZ GE} e
KILOMETRES Ke)
ue r
ae ¢
ae .

A Barmouth of :

—_— 0 s

wie
. = ie Pace “2180 SARNS
CON ries BA — BADRIG
WAT Fr A Ani BW — BWCH
ee Ve 0 WA - WALLOG

Fic. 18. Distribution of lithic arenite fragments. Contour interval 10%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 67

westward from the other part of Sarn Wallog. Within this zone, which is delimited
by 5 % values, there are several stations with values above 10%, namely 109 (120%),
201 (14:0%) and 19r (14:9%). These deposits are poorly sorted, and it is suggested
that they have been derived by erosion of the outer parts of Sarn Wallog and trans-
ported northwestward by bottom currents.

In the northern part, a linear band of slate-containing sediments commences in
the vicinity of a coastal source near Tal-y-garreg, and extends in a northwesterly
direction until it reaches Sarn Badrig. Additional slate fragments in this northern
zone may originate south of Sarn Bwch near Towyn and, perhaps, be joined by
others eroded from the cliffs north of Sarn Bwch. There are no samples with
significantly higher values, i.e., at least 10%, in the northern region which would
indicate that slate fragments are being introduced from beyond Sarn Badrig. How-
ever, stations around the seaward end of Sarn Badrig (all over 5 °%) may be influenced
by lithic grains eroded from the sarn itself. From the end of Sarn Bwch westward,
a zone of coarse sediments with slate fragments in amounts above 5 % suggests that
these are derived from Sarn Bwch. Station 196 with 22-:0% is anomalously high,
and it may reflect textural control (6 Mz, — 1-2).

Locally within the Mawddach estuary, some deposits exceed 5 % slate fragments,
but these are restricted to areas of known temporary entrapment and do not suggest
that slate fragments are being contributed to the offshore sands by any source
related to that estuary. This observation is further supported by the absence of a
dispersal shadow seaward of the estuary mouth. While the Dyfi does not appear
to be a transport zone carrying slate particles to the bay, a minor enrichment zone
is apparent in the upper estuary (163 with 25-9 %, 169 with 11-5 %, 158 with 10-0 %)
Slightly higher values at Stations 146 and 167 are indications of the coarser grained
deposits there. In summary, the two important dispersal centres are the cliffs at
Borth, and the seaward end of Sarn Wallog. Minor amounts of slate are deposited
in the upper Dyfi estuary.

Crystalline rock fragments. All crystalline rock fragments, regardless of their
type or clan, are grouped into a single classification in Text-fig. 19. Three dispersal
zones are recognized. First, a zone delimited by a 2% contour, which extends
irregularly westward just north of Sarn Wallog. Crystalline rock grains in this
area are of several varieties. Second, a zone in the northwestern part of the
survey wherein the crystalline rock fragments are also of several varieties. This
area appears to be related to a dispersal centre beyond the limits of the survey.
Third, inshore, in the northern part of the area is a dispersal zone extending from
just north of Sarn Bwch to the vicinity of Eglwys Llanaber. A typical dispersal
shadow is established adjacent to the coast between Sarn Bwch and the entrance
to the Mawddach estuary.

While the upper Mawddach estuary contains crystalline rock fragments in amounts
exceeding 3%, indeed as much as 28% in the coarse-grained stream deposits at
Dolgellau, there is no suggestion that the Mawddach delivers any significant amount
of crystalline rock fragments to the bay. In fact, the middle Mawddach estuary
area is noticeably deficient in crystalline fragments, an observation which is sup-

MINER. 2, 2. 5

68 SEDIMENTATION IN NORTHERN CARDIGAN BAY

On 15 2 84) ISeiGea7 SEs Om
SS = SS SS a Se S|

KILOMETRES

= 2/029
125 ;
Barm

Chee 2140

. A) oy) Aberdyfi ee

SARNS
BA — BADRIG
BW —- BWCH

WA — WALLOG

|

Fic. 19. Distribution of crystalline rock fragments (all varieties). Contour interval 2:0%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 69

ported by the data for Stations 129 (06%), 140 (0:9%), 141 (0°6%) and 125 (1:8 %).
If crystalline fragments in significant amounts are presently transported through the
Mawddach estuary, they would have been observed in these samples. Certain of
the crystalline fragments found in offshore sands west of Ty-wen resemble, in thin-
section, crystalline rock types occurring in the Cader Idris area nearby.

Chert. It is apparent from the chart (Text-fig. 20) that these grains are being
dispersed westward from the vicinity of Borth, and that their source is the Aberyst-
wyth grit beds exposed in the cliffs nearby. Since detrital chert fragments are
normally derived from eroded limestone, this suggestion of local origin, i.e., the
Aberystwyth grits, implies a complex relationship of multi-cycled grains, first
deposited in the Aberystwyth grits and now being eroded and deposited in the bay.
A small development offshore from Towyn suggests that chert fragments in the
deposits there are derived from the exposed glacial sediments along the shore.
There is no evidence that chert grains are transported to this area by the rivers,
or by currents from beyond Sarn Badrig.

(c) Comparative petrography

Use of photomicrographs taken of thin sections for certain of the Cardigan Bay
and associated stream samples provides a direct, visual method for sediment com-
parisons. While photomicrographs taken at low magnifications are not suitable
for classifying specific quartz varieties, they are, on the other hand, useful for
illustrating major compositional variations between selected samples.

One of the initial questions in this study involved the determination of detrital
types which might be carried by the rivers and deposited in the estuaries, or in the
bay itself. Certainly the preponderance of fine grained quartz sand in the Dyfi,
as well as the Mawddach estuary, should be related to the transport of similar quartz
sand in the adjacent rivers if the rivers supply the sand. However, thin-sections
of river sediments upstream from the Dyfi estuary are all but devoid of fine and
medium size quartz grains. Even such samples as the one at Station 160 in the
Dyfi, a sample containing sand sized lithic material, is noticeably lacking in quartz
grains as discrete clastics. To illustrate the major change between sediments in
the Dyfi river and those in the upper Dyfi estuary, consideration will be given to
four typical photomicrographs. In Pl. 3, fig. 1, a photomicrograph of the river
sediments at Station 162 near Machynlleth, there is a predominance of lithic frag-
ments. This photomicrograph is, indeed, typical of most of the mobile sediments
in the Dyfi river in the vicinity of Machynlleth. While rare fragments of arenites
have been seen in the course of field study, most of the clastics are fragments of
slate and shale. If mechanical breakdown of lithic fragments being rolled or
otherwise transported by the river was the source of sand size quartz, i.e., by mech-
anical disaggregation, then it would be expected that parent lithic fragments should
contain abundant sand size quartz grains. Pl. 3, fig. 1 shows that this is clearly
not the case. In Pl. 3, fig. 2, a photomicrograph of a thin section of sample 161
some 3 km. downstream, the lithic fragments are slightly smaller in size but are the
same in composition. It will be noticed that the only real change is in the texture

79

SEDIMENTATION IN NORTHERN CARDIGAN BAY

Ou 2223) 45 66E 7B Ont
KILOMETRES

BA — BADRIG
BW — BWCH
WA — WALLOG

zis

Aberystwyth

Fic. 20. Distribution of chert fragments.

———— |

Contour interval 1:0%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 71

of the sediments. The lithic composition remains that of slate. However, Pl. 4,
fig. 1, for Station 159 in the lower Dyfi valley or upper Dyfi estuary, shows a pro-
nounced change in the aggregate composition of the sediment. This represents a
downstream transport direction of about three more kilometres. At this station
there is a mixture of lithic fragments similar to those observed at upstream stations
and fine quartz sand. At Station 159, the texture is such that the sediment sample
is a bimodal one, i.e., large lithic fragments and small quartz grains. Station 159,
incidentally, is within the transition environment of “ upper estuary/lower river ”
sedimentation. Farther downstream (Pl. 4, fig. 2, St. 163), the entire textural suite
is of sand size and is obviously quartzose in composition. It may be concluded
from these four petrographic examples that the fine grained quartzose sand is
restricted in its deposition, and does not extend beyond the influence of the highest
tides. Consequently, it would seem that while lithic material, normally slate, is
being gradually moved downstream, there is no suggestion that quartz sand, such
as is found in the estuary, is being moved likewise. Any phyllosilicate grains which
might be eroded from larger lithic fragments during the course of stream transport
are probably flushed from the estuary, or in some cases entrapped within the marsh
(grass stabilized) deposits along the south side of the Dyfi estuary. These con-
clusions are supported by X-ray diffraction data. At Stations 162, 161, 159 and
163 the concentrations of phyllosilicates are respectively 57, 57, 33 and 14%. This
decrease of total phyllosilicates downstream is in agreement with petrographic
observations.

Another example illustrating the use of comparative petrography is the characteri-
zation of sediments near the cliffs at Borth, and their comparison with deposits
further offshore. Pl. 5, figs. 1, 2, photomicrographs of samples from Stations
148 and 144 from the littoral zone at the base of the cliffs south of Borth, show an
entirely different grain type assemblage from the Dyfi system. Here the lithic
fragments are predominantely arenites and are mixed with quartz grains. In
Pl. 6, fig. x (St. 150), a sample from the littoral zone at the end of the cliff section
associated with the beach environment at Borth, it may be seen that the amount
of lithic fragments is somewhat decreased, but with a concomitant increase in quartz
grains. The deposit is also one of finer size than the littoral deposits at the base
of the cliffs to the south of it. Such a relationship suggests that within the few
kilometres of northward longshore drift transport, there is a diminution of lithic
material, and this may be explained by the mechanical breakdown of the softer
rock fragments into their constituent grains. If we compare, then, these sediments
near the cliffs with some deposited farther offshore (PI. 6, fig. 2), it is readily apparent
that the bay sands seaward are relatively richer in quartz grains than are the littoral
deposits adjacent to the cliffs. The photomicrograph of a sample from Station 7
(Pl. 6, fig. 2) is a typical example of the fine grained sand flooring much of northern
Cardigan Bay.

In summary, gross petrographic comparisons agree with the findings based on
single grain studies and suggest that fine quartz is not being delivered to the estuaries
by the adjacent rivers, but that it is being contributed to the bay and, in turn, the
estuary sediments by active erosion of beds exposed along the shore.

72 SEDIMENTATION IN NORTHERN CARDIGAN BAY

VI. X-RAY MINERALOGY

Although X-ray diffraction analyses of marine sediments, using the powder method,
were made as early as the late 1920’s (Correns, 1935), it was not until after World
War II and the advent of modern diffraction apparatus for making rapid repro-
ducible analyses that sedimentologists began reporting the dominant mineral
composition of marine clastics (Grim, 1953). Many X-ray studies of the past few
years have been concerned only with differentiating the clay types. However,
gross composition data for quartz, feldspars and phyllosilicates are equally valuable
for classifying and comparing Recent sediments with other marine deposits, and for
providing a reference for chemical and petrographic data. There are, however,
several limitations inherent in the X-ray powder method: (1) non-crystalline
components are not recorded—volcanic glass, for example, would be overlooked if
diffraction analysis were used without recourse to thin section study; (2) gross
analysis does not provide quantitative data for minerals present in quantities below
about 1% ; and (3) the method cannot differentiate the textural states of the several
components.

(a) Distribution of minerals

Muscovite. Mica occurring in the Cardigan Bay deposits is in the form of well
crystallized dioctahedral muscovite and is the only one of the r0A minerals present.
With the exception of some very small quantities present as clay-size clastics in
estuarine marsh deposits (less than 1%), the mineral occurs as a constituent of lithic
grains, chiefly slate and arenite fragments.

From a study of the data, it is seen that the major fine sand area offshore contains
less than 5% muscovite. The average, in fact, is close to 3% both for the fine
sand area north of Sarn Bwch, and for the fine sand area off the Borth/Towyn coast.
Fine sands along the south side of the Dyfi estuary, particularly in the marsh
environment, average about 6% 10 A mica, which, from petrographic evidence, is
contained in very small lithic fragments, normally the smallest sizes of slate frag-
ments readily discernible. Station 146 with 5 % mica is a “‘ coarse sand ” (0-3 ¢ Mz),
and Station 167 (13 %) is a “ very coarse sand’ (—o-7 ¢ Mz) ; in both deposits the
higher proportion of mica is due to the larger percentage of coarse lithic fragments
present. The Afon Dyfi channel deposits (Text-fig. 2) just upstream from the head
of the estuary (Stations 159, 160, 161 & 162) are sands, excepting Station 162 (near
the town of Machynlleth), which is gravel (—2:3¢ Mz). Samples from these
stations are reported to have 12, 17, 22 and 23% mica content respectively, all
attributed to lithic fragments.

Stations 159 and 163 in the transition area of the uppermost estuary and lowermost
stream influences are interesting and deserve special attention. Station 163, an
uppermost estuary deposit, is a fine sand (2:9 ¢ Mz) ; it texturally resembles the
other “‘ typical ’’ estuary fine sands and possesses a 10A mica content of 5%. Yet,
Station 159, a very fine sand (3:0 ¢ Mz), comparable in grain size to Station 163
and located about 1 km. farther up the channel, has a mica content of 12%, a pro-
nounced difference. In the same general area, Stations 164 and 164A, very coarse

SEDIMENTATION IN NORTHERN CARDIGAN BAY 73

sands (—o-7 and —o-8 ¢ Mz, respectively) have 25 and 24% mica content each.
Thus, a significant change in mica content occurs between the uppermost “ typical ”’
fine sands of the estuary and the deposits of the lower Afon Dyfi and its smaller
tributaries. Fine sand in the lower river and fine sand in the upper estuary are
texturally similar, yet there is a three-fold decrease of mica in the estuary deposit.
In fact, most fine-grained sands in the main tidal channels on the north side of the
Dyfi estuary are depleted in mica, the average content being only 2%.

The littoral deposits extending southward along the cliffs near Borth all have
mica exceeding 10%, except a beach deposit containing only 6% mica. There
is, however, much less mica in the more seaward deposits. These fine sands im-
mediately offshore average only about 3%. As mentioned previously for texture,
a sharp boundary exists between the immediate nearshore coarse-grained deposits
and the offshore fine sands.

In the Mawddach estuary, the fine and very fine sands constitute, areally at least,
the major portion of the deposits encountered there, and contain between 2 and 6%
mica. Silt deposits in shallow, protected, partly marsh areas have slightly higher
mica contents (11% and 7%). Beyond the mouth of the Mawddach estuary (Text-
fig. 2), just off Barmouth, inshore Stations 92, 93, 94, 122 and 135, all fine sands,
have about 2% mica content.

The coarse and medium sand deposits, offshore and northwest of Towyn, were
found to contain more mica than the fine sands surrounding them ; similarly, the
tongue of poorly sorted coarse and very coarse sands extending northwestward from
the outer Sarn Wallog area has higher mica percentages than the surrounding fine
sands, an increase due to lithic fragments.

In summary, X-ray diffraction data reported for these deposits suggest that the
muscovite content of the fine sands is distributed uniformly and in small amounts,
usually 2 to 3% with slightly more in the estuarine fine sands. Coarser, poorly
sorted deposits, which are bimodal mixtures of sand and fine gravel, exhibit signi-
ficant enrichment in mica. The stream deposits, regardless of their specific textural
class, though they are usually coarse grained, also possess larger amounts of mica
than do the offshore sands. Moreover, the transition from deposits containing little
mica in the bay and estuaries, to the stream and eroded shore deposits is a sharp
one. A similar sharp change exists between the well sorted fine sands and poorly
sorted, coarser deposits in the bay proper.

Chlorite. The only other phyllosilicate, or ‘‘ clay’ structure mineral, present in
quantities above the X-ray diffraction limit for gross analysis is chlorite. Although
minor lattice variations in this 14 A mineral do occur, the chlorite present in Cardigan
Bay deposits is, crystallographically at least, the same throughout the area. Such
a conclusion is based on its oor (14:12 A) and 003 (4-71 A) reflections, and 002
(7-06 A) and 004 (3°53 A) reflections.

Major fine sand deposits offshore, excepting the 2 km. wide zone off the Borth
beach, contain less than 5% chlorite and have a mean of about 39%. There are no
pronounced trends within these find sand deposits as chlorite seems to be a ubiquitous
component. Increased percentages of chlorite are reported for a zone extending

74 SEDIMENTATION IN NORTHERN CARDIGAN BAY

about 2 km. offshore, and parallel to the coast from Borth northward to about the
mouth of the estuary. Sand in this relatively restricted area (Text-fig. 2) contains
amounts of chlorite greater than 5%, but less than 10%, e.g., at Station 25, a very
fine (3-1 ¢ Mz), well sorted (1-13 So) sand contained 8%. Chlorite progressively
decreases in littoral sands with increasing distance northward from the contact
of the Borth beach with the eroded cliffs. These beach samples, except 152
(@ Mz 2-1), are medium-grained sands. Close inshore samples near the cliffs
between Borth and Sarn Wallog (poorly sorted deposits containing coarse lithic
fragments) contain abundant chlorite. Also in the southern part of this survey,
a tongue-shaped zone of sediments with high chloritic content is delimited for much
the same stations as those previously shown to have abundant mica present.

Excepting Stations 158A (3%) and 168 (3%), all samples from the Dyfi estuary
contain chlorite in excess of 5%. Fine grained marsh deposits along the southern
side of the estuary contain significantly larger amounts of chlorite than do other
sediments of similar texture in the estuary (a similar situation exists for the 10oA
mica distribution). Samples from the transition zone of sedimentation in the upper
estuary/lower stream channel region provide evidence of an abrupt decrease of
chlorite. Station 163, in the uppermost estuary, with 9% chlorite, and Station 160,
in the lowermost stream channel, with 35°%% chlorite, exhibit a marked difference
in chlorite content, even though both samples are fine sands (¢ Mz 2-9) and both
are well sorted (So: 1-12 and 1:20). Still farther upstream, samples collected in
the vicinity of Machynlleth were found to contain about three times more chlorite
than the uppermost estuary sands. All coarse grained deposits in the tributary
streams on the south side of the estuary contain more chlorite than do the estuary
deposits.

Deposits in the Mawddach estuary contain slightly more chlorite than do the
nearby offshore fine sands. With one exception, which contained 4°% chlorite, all
samples in this estuary are reported to have chlorite in amounts exceeding 5%.
Three local areas in the estuary, represented by Stations 136, 140 and 126, are en-
riched in chlorite, with rz, 1r and 13 % respectively.

In general, chlorite is found only sparingly in the fine sands flooring the bay and
is usually present in amounts less than 5%. It is slightly enriched in the fine
sands off the Borth beach, and in the fine grained deposits of both estuaries.

Combined phyllosilicate suite. Charting of the combined mica and chlorite data
is instructive, both in further definition of the dispersal pattern in the bay and in
providing a clay mineral framework for comparison with the charted chemical data.
Although the dispersal pattern for the combined phyllosilicate group is much the
same as for the individual mica and chlorite species, it presents, nevertheless, a
substantially more detailed distribution picture. Whereas the chlorite data delimit
a facies off Borth extending northward only some 2 or 3 km. seaward, the combined
10 A/14 A suite shows a composite facies extension considerably farther seaward.
Likewise, the tongue-shaped pattern of sands extending northward off Sarn Wallog
are critically delimited. As expected, the Dyfi and Mawddach estuarine deposits
show higher contents of platy minerals than do the fine sands offshore. Moreover,

SEDIMENTATION IN NORTHERN CARDIGAN BAY 75

river channel deposits are much enriched, as are sediments associated with coastal
and sarn erosion.

Plagioclase. The presence of Na-plagioclase was established for many of the bay
deposits (Text-fig. 21). Plagioclase, as well as orthoclase, occurred frequently as
very small, sub-rounded grains, and because of this textural relationship, i.e., smaller
in size than its host quartz, the use of gross X-ray diffraction analysis for establishing
feldspar abundance is preferred. For most of the fine sands in the bay, the plagio-
clase content is less than 5 94, the mean being between 3 and 4%. A zone of deposits
containing over 5 % plagioclase extends northward from the base of the cliffs between
Borth and Sarn Wallog. Inasmuch as the deposits in this particular area range in
size from very coarse “‘sands”’ (near the base of the cliffs) to very fine and fine
sands (offshore), plagioclase distribution here may not be totally size dependent ;
indeed, some plagioclase is known to be contained in rock fragments.

Excepting Stations 154 (4%), 158A (4%) and 165 (2%), sediments in the Dyfi
estuary contain plagioclase in excess of 5°%. Unlike the major contrast in the
phyllosilicate abundance between estuarine and stream deposits, the upper estuarine
deposits (Station 169, 6°) contain much the same amount of plagioclase as do the
Dyfi stream sediments and nearby tributaries. Bottom sediments northwest of
Towyn also exhibit a slight increase in plagioclase content. Similarly, off the
eroding cliffs northwest of the Tal-y-garreg beacon, plagioclase increases slightly
(Text-fig. 21). ;

Mawddach estuary deposits, like those in the Dyfi, contain more plagioclase than
do the immediate offshore sands. A slightly enriched zone extending from the
mouth of the Mawddach estuary seaward in a southwesterly direction (Stations 93,
92, 91 and 41) suggests that some plagioclase might be in transport there. An
irregularly outlined zone of sediment containing plagioclase in excess of 5% occurs
adjacent to Sarn Badrig. The increase in plagioclase in this area is related to active
erosion of the sarn and to a source farther north, perhaps in the Tremadoc Bay area.
One sample with 20° plagioclase content was collected from the stream bed at
Dolgellau. The abundant plagioclase found there does not seem to be related to
the overall pattern, since deposits in the uppermost Mawddach estuary are not
proportionately enriched. Station 137, however, does contain 28 % crystalline rock
fragments.

In summary, plagioclase content of the bay sands is less than 5 °% at most stations,
except off Borth and adjacent to Sarn Badrig (5 to 10%). Estuarine and stream
deposits show plagioclase values between 5 and 9%.

Orthoclase. Text-fig. 22 shows that there are two regions of modest enrichment
of orthoclase in the bay. A bipartite area of sediments having over 4% orthoclase
extends southward from off Towyn to just north of Sarn Wallog. Accumulation
of orthoclase in these sediments may be controlled, in part, by the texture of the
host fraction, as the finest sands are abundant here. Farther seaward, orthoclase
is randomly dispersed. The only other area in the bay where orthoclase is present
in quantities of 4% or more is the elbow-shaped zone extending from off Ty-wen
northward to the vicinity of Station 120, thence southwesterly along the edge of

76 SEDIMENTATION IN NORTHERN CARDIGAN BAY

7 Se
Oe Pee SG 9 EN Osh) eel eS Rae
—— iT Ft ——————— ot ———

KILOMETRES 2 . 4

Mawddach
Estuary.

Wwf 2 9 lal ~yPgarreg

te
e £ ens ; e
e eS e e ‘

\ e ° 2 a
° e e \
e eo : 3
pees AK BA — BADRIG
ON ES BW - BWCH
a re
7 ° WAT NO el i * A WA — WALLOG
e J!
e
5 4 : e 57

Fic. 21. Distribution of Na-rich plagioclase based on X-ray data.

Contour interval 5:0%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 77

Sarn Badrig. Here again, K-feldspar dispersal may, in part, be texturally controlled,
and more related to fineness of the sands (Text-fig. 3) than to localized introduction
from some nearby source.

Only two samples from the Dyfi estuary were found to contain as much as 4%
orthoclase, and none of the stream deposits in the Dyfi province contained over
3% K-feldspar. A similar situation exists for the Mawddach estuary and stream
deposits.

Quartz. Diffraction data for total quartz (the dominant mineral) are reported,
but since the distribution of its petrographic varieties has been discussed in detail
earlier in this report, it is not reviewed in this section.

Calcite. Occurring predominantly as bioclastic debris, calcite is rather evenly
distributed in the fine sands of the bay, for which values between 5 and 10% are
reported. The only deposits containing less than 5°% calcite are: (1) the trans-
littoral neritic sands along and within a kilometre or two of the coast, (2) an elongate
area about 10 km. west of Aberdyfi, and (3) a tongue-shaped zone extending west-
ward from Borth. Although Scrobicularia shells are found in some of the estuary
deposits, no noticeable concentration of calcite is observed there. On the south
side of the Dyfi estuary, slightly higher values for calcite are reported than for
deposits on the north side. Calcite was not found in any of the stream deposits
associated with either estuary. Sediments with calcite in excess of 10% are found
north of the end of Sarn Wallog and are associated with poorly sorted, coarse sands.
The increase there is attributed to relatively abundant shell fragments.

Other Minerals. For a very few samples, faint traces of peaks for pyrite and
hornblende were noted on the diffractograms. It is estimated they would represent
1% or less of the sample. Petrographic study shows that these are related to rock
fragments. A very faint indication of the 2-89A peak for dolomite was found on
many diffractograms. Dilute HCl treatment was given to a number of the samples,
and the 2:89A peak disappeared on subsequent analysis, confirming the carbonate
relationship. Dolomite data are reported as 1%, since abundance values cannot be
determined below this limit. (See appended data for Aberystwyth grits samples
in reference to the probable source of detrital dolomite.)

(b) Phyllostlicate relationships

A plot of chlorite and mica data (Text-fig. 23) shows that a common relationship
exists between these two phyllosilicates. In fact, an approximate 2 : 1 chlorite/
mica ratio is common for many bay deposits, regardless of their texture or their
location. Although the diffraction data were plotted one with another, for the
various minerals, chlorite and mica were the only two minerals which showed a
clearly defined regression. Petrographic evidence confirms this finding.

It is clear from Text-fig. 23 that the reference rocks (Table I) from mid-Wales plot
reasonably close to the regression trend for the bay sediments. Some scatter is
seen, however, for several of the rocks. It is not likely, nor would it be reasonable
to suggest, that 26 reference rocks represent proportionate coverage of the proven-

78

Ont

Fic, 22,

SEDIMENTATION IN NORTHERN CARDIGAN BAY

20364 S67 691014
KILOMETRES

Za
a“

ae ee Mawddach ~
\y ie Estuary: wy

SARNS
BA — BADRIG
BW- BWCH
WA - WALLOG

Distribution of orthoclase based on X-ray data. Contour interval 4:0%.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 79

ance. However, if the various rocks with both greater and lesser quantities of mica
(relative to the regression) were to be eroded equally, and equally mixed in the bay
with the others, the average ratio of the detritus would be very near that established
by the regression.

CHLORITE %
25 30 35 40 45

Fic. 23. Correlation of chlorite with muscovite for Cardigan Bay sediments (dots) and
for reference rocks (crosses) from mid-Wales. Note the linear relationship for the bay
sediments and many of the reference rocks. See text for discussion.

Although the mica-chlorite plot shows that a few sediments in excess of 20%
mica are displaced from the trend, and might be interpreted as the result of lattice
variation resulting from diagenesis after entering the bay, the author does not
believe that the data are sufficient in number to recognize such alteration. In short,
the chlorite/muscovite ratio for combined clastics provides additional support for
the premise that these sediments are largely of local derivation.

VIESDIST RILEY TION OF ELEMENTS

Although the interplay of several factors such as the relative proportion of detrital
material, variability in the composition of the argillaceous material, and the sulphide,

80 SEDIMENTATION IN NORTHERN CARDIGAN BAY

heavy mineral and organic content affect the trace element content of sediments
(Carr & Turekian, 1961, p. 42), it is nevertheless possible, by charting the data for
the several elements, to obtain a reasonable interpretation for their distribution.
Vectors have been drawn on the charts so that a composite chart (Text-fig. 34)
could be prepared, and sources as well as areas of enrichment have been noted.

Aluminium. Extending northwestward and in part landward from the seaward
extremity of Sarn Wallog is a wide tongue-shaped zone of bottom sediments which
contain alumina in excess of 4%. This correlates with the distribution pattern of
coarse clastics (Text-fig. 3), and establishes a lithic fragment/alumina relation. A
narrow zone of sands with alumina in excess of 4%, related both to feldspar and to
lithic fragments, extends northward from the cliffs at Borth. Between Towyn and
Sarn Bwch, alumina-bearing sediments reflect the known dispersal of lithic material
in that area. Additional evidence that alumina is related to the dispersal of rock
fragments is found in the western mid-bay region, where the higher alumina content
is related to the lithic fraction of the sediments.

Barium. This is a normal constituent of feldspars and phyllosilicates, and is
preferentially distributed in Cardigan Bay. In the southern part of the survey
(Text-fig. 24), a distribution pattern of Ba (over 100 ppm) suggests that the cliffs
at Borth and the outer edge of Sarn Wallog are sources of Ba-bearing detritus. In
the northern part of the survey, Ba rich sediments are derived from the glacial
debris exposed near Ty-wen, and some are transported in a northeast trend zone.
The latter are related to the bathymetric salient (Text-fig. 2) already established
there. Although both river deposits and marsh deposits are enriched in Ba, there
is no apparent transport of the Ba-rich sediments seaward from the estuaries.
Moreover, most well sorted beach sands are low in Ba, relative to offshore deposits.

Boron. This element is present in all analysed samples ; with one exception,
it occurs in amounts exceeding 20 ppm. Furthermore, the data suggest that there
are two definite zones of boron enrichment. The distribution of these two areas
(Text-fig. 25) is such that concentration of boron-rich sediments occurs relatively
near the coast in fine sands, and the most pronounced of these two zones is that area
encompassed by the 40 ppm contour from near Borth northward to Sarn Bwch.
In the northern part of the survey, an important distribution area extends from
near the coast southwestward to the terminus of Sarn Bwch. Within the Mawddach
estuary, several stations exceed 40 ppm, but no systematic pattern of distribution
is recognized. For the Dyfi estuary, there are two zones with sediments containing
greater than 40 ppm: one on the north side of the estuary and one on the south
side. Both are in marsh sands. Neither relates to dispersal centres.

Surprisingly, samples collected from the several rivers (with many lithic grains)
do not show that boron is significantly concentrated in stream deposits, e.g., a river
sample at Dolgellau contained 24 p.p.m., and a sample associated with the Dyfi
drainage area contained 82 p.p.m., these being the two extremes for river sediments.
In summary, boron, primarily in tourmaline, is concentrated in the fine sands
offshore between Borth and Sarn Bwch, and in a narrow zone in the northern part of

SEDIMENTATION IN NORTHERN CARDIGAN BAY 81

(OSE 2, ss ey ates WS Fy Le eh |
a
KILOMETRES

i

BA BADRIG
BW BWCH
WA WALLOG

Fic. 24. Distribution of barium based on spectrographic data. Contour interval
100 ppm.

82 SEDIMENTATION IN NORTHERN CARDIGAN BAY

OER 2S SS MARES EG M/ass eon OMil

KILOMETRES

Mawddach
Estuary

\Bar mouth’
(Xe)

BA — BADRIG
BW — BWCH
WA — WALLOG

a

Fic. 25. Distribution of boron based on spectrographic data. Contour interval 20 ppm.

C Aberystwyth

SEDIMENTATION IN NORTHERN CARDIGAN BAY 83

the survey. It does not appear to be related to a source associated with either the
Afon Dyfi or the Afon Mawddach.

Calcium. Although some calcium is present in phyllosilicates and feldspars in
trace amounts, the major portion is contained in bioclastic debris and, as such,
indicates the proportionate enrichment of calcite in marine sands. Within the
present survey, the amount of calcium present at any one station in the bay varies
between 1-32% near Eglwys Llanaber and 9:00% some 18 km. west of Towyn.
However, most of the bay sands contain less than 3 % calcium and only three zones
of enrichment are distributively important. These are: (1) a band of sediment
containing more than 3% extending along the south side of Sarn Badrig, (2) an
elongate zone extending northwest from the end of Sarn Wallog which correlates
with increasing grain size and poor sorting, and (3) an area of calcium enrichment
associated with coarse deposits and shell debris in the western margin of the survey.
In general, calcium decreases toward the headward end of each estuary ; otherwise,
there is no systematic distribution within these smaller bodies of water. Moreover,
littoral sands from just north of Towyn in the vicinity of Station 174 (1-80%)
northward to Sarn Badrig are noticeably deficient relative to the offshore deposits
in that they contain less than 2° calcium. Obviously, it is not useful as an indicator
of provenance for these bay sands, since carbonates are not found nearby.

Cobalt. With the exception of those deposits associated with coastal and sarn
erosion, cobalt is not concentrated in these marine sediments. In fact, for the
region as a whole, there is little variation between the various offshore stations, and
no cobalt dispersal pattern is reported. The contrast in distribution of cobalt
between river sediments and nearby estuary sediments should be noted, however.
Cobalt values for river samples associated with the Mawddach are considerably
higher than the associated estuary sediments. The same is true for the Dyfi. Thus,
estuary and river sediments are apparently unrelated.

Chromium. The major zone of chromium enrichment extends from the cliff
area south of Borth northward, terminating some 4 km. west of Aberdyfi (Text-fig.
26). At this point, a narrow zone of chromium-rich sediments is developed south-
westward. Furthermore, fine sands with high chromium values are transported
across Sarn Bwch, and no interruption of the distribution plan occurs in the vicinity
of the sarn. Another major distribution within the southern half of this survey
occurs as a belt extending northward from the outer end of Sarn Wallog. This
pattern follows a zone of coarse grained, poorly sorted detritus (Text-fig. 6). Only
one distribution area of importance is noted in the northern half of the survey.
This, an elongate, narrow zone extending diagonally across the inshore fine sands,
is defined by Station 237 (87 ppm) at one extremity, and Station ror (83 ppm)
at the other.

No evidence is available to suggest that chromium distribution is associated with
the Mawddach and Dyfi estuaries, although slight enrichment in the marsh deposits
is reported. On the average, a two- to three-fold increase over the offshore sands
is noted for river deposits. In conclusion, sediments enriched in chromium (over
50 ppm) are restricted, as major distribution areas, to a narrow mid-bay belt in

MINER. 2, 2. 6

SEDIMENTATION IN NORTHERN CARDIGAN BAY

84

Ot 22123) «4,5 ie Oe BES 10 ah}
es ss Ss ——_
KILOMETRES

Mawddach
Estuary

° 450
Barmouth of:

32> 56)

SARNS
BA — BADRIG

BW — BWCH
WA — WALLOG

\
‘ sel
. / Aberystwyth

Contour interval

Fic. 26. Distribution of chromium based on spectrographic data.
50 ppm.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 85

ORM Cn Sa nAstS Gan 22 BLS JOLT
KILOMETRES

Mawddach
Estuary . 3
eee
Barmouth PY

BA — BADRIG
BW —- BWCH
WA — WALLOG

‘ berystwyth

Fic. 27. Distribution of copper based on spectrographic data. Contour interval Io ppm.

86 SEDIMENTATION IN NORTHERN CARDIGAN BAY

the north part of the survey, to a narrow belt extending north from Sarn Wallog,
and to an irregularly shaped zone extending northwestward from the cliffs at Borth.

Copper. One of the most rewarding of the several charted element distributions
is that for copper (Text-fig. 27). The distribution of sediments enriched in copper,
i.e., Over 10 ppm, is basic to understanding the distribution of nearshore sands
within the survey and to locating those areas offshore with sands which owe their
origin to local provenance.

A definite distribution is charted for three lobe-shaped zones extending seaward
from the cliffs near Borth. Within this distribution area several samples are reported
with values exceeding 20ppm Cu. Likewise, littoral samples along the coast
between Sarn Wallog and Borth are high in copper. Copper enrichment of these
nearshore deposits is related to the cliff exposures south of Borth. Anomalously
high values are reported for the coarse grained, poorly sorted sediments extending
northwestward from the end of Sarn Wallog.

A coastal zone extending from Towyn northward and beyond Sarn Bwch contains
sediments enriched in copper. In fact, some of the highest reported values are for
stations within this nearshore area, obviously related to the eroding cliffs of glacial
debris at Towyn. In the northern part of the survey, south of the Sarn Badrig
shoal near Mochras, several stations are charted with values above 20 ppm; these,
as well as the general distribution of samples with less than Io ppm in the far northern
part of the bay, strongly suggest a distribution related to a source or sources beyond
the limits of this survey, perhaps in Tremadoc Bay. In the Mawddach and Dyfi
estuaries, several zones of enrichment occur locally, but there is no trend suggesting
sources upstream.

Ivon. It must be emphasized that comments on distribution relate to the total
content of iron (as Fe,O,) in the sediments, regardless of its mineralic host. Iron
remains rather constant in amount throughout most of the sand areas, being between
2 and 3%; however, it is modestly enriched in three rather limited zones. The most
pronounced of these extends northward from the outer edge of Sarn Wallog. Another
area of enrichment (over 3 °%) is that which parallels the cliffs near Borth and extends
northward to just off the estuary mouth. Here again, a progressive decrease is
noted for those stations farther from the littoral zone south of Borth. The only
other area of enrichment is a small zone just northwest of Towyn. In the Dyfi
estuary, iron is preferentially distributed along the south shore, but in the Mawddach
estuary there does not seem to be any definite distribution plan. Considering both
estuaries, neither seems to be related to sources, or to transport of the iron-bearing
minerals found offshore.

Gallium. Throughout most of the study area, gallium is rather evenly distributed
in amounts between 2:0 and 5:0 ppm; however, in the southern part, two zones
are present where the sediments contain 5:0 ppm, or more, gallium. The first of
these is that which extends northwestward from the end of Sarn Wallog. This
zone shows a progressive decrease in the amount of gallium reported at stations
increasing with distance from the sarn. These stations lie within a belt of coarse
grained “‘ sands’, which extend away from Sarn Badrig (Text-fig. 6), and the two

SEDIMENTATION IN NORTHERN CARDIGAN BAY 87

distributions are related. The second area is developed as an elongate distribution
of gallium-enriched sediments extending northward from the cliffs near Borth, and
paralleling the nearby beach as far north as the estuary mouth.

Deposits in the Mawddach estuary do not show any recognizable distribution plan,
even though certain of the samples contained gallium in amounts exceeding 10 ppm.
On the other hand, sediments in the Dyfi estuary show enrichment of gallium in
the marsh along the south side of the estuary. River sediments, even in small
streams, reflect their high content of lithic fragments by pronounced increase in
gallium.

Potassium. While potassium is present in amounts exceeding 1% in the river
deposits, it is noticeably decreased in amount in the offshore and estuary sands,
and only eight samples were reported to contain amounts exceeding the above value.
For the survey as a whole, particularly in the fine sands, potassium is distributed
in amounts less than 0:-5°%. Nevertheless, three zones are noteworthy, namely :
the zone extending northwestward from the end of Sarn Wallog, the narrow coastal
zone extending northward from the cliff area south of Borth, and the small develop-
ment between Towyn and Sarn Bwch.

Magnesium. The distribution of magnesium in amounts exceeding 0:5% is
limited to 15 stations in the entire offshore part of the survey. Of these, only those
in the littoral zone at the base of the cliffs south of Borth, and in the zone of coarser
sediments northwest of the seaward tip of Sarn Wallog are of importance. In both
of these areas, the enhancement of magnesium is due to higher phyllosilicate content.
In the Dyfi estuary, there are only two stations with magnesium reported in excess
of 0-5 %, and such limited control cannot be expected to establish any meaningful
pattern there. The situation is much the same for the Mawddach estuary. In
short, magnesium is present in all samples, but for most of the bay sands it shows
little variation. Due to their higher lithic content, the river deposits are about
three times richer in magnesium.

Manganese. The distribution of manganese (Text-fig. 28) is complex and, as
such, necessitates careful consideration of its pattern of enrichment, particularly
in the sands. Manganese is considerably enriched in these marine sands in compari-
son with certain other areas. For example, Moore (1963) reported maximum
values of about 300 ppm Mn for Buzzards Bay deposits which contained much
higher amounts of phyllosilicates. In the present study, however, even well sorted,
fine grained, non-argillaceous sands include manganese in excess of 500 ppm.

In the southern half of the survey, two prominent areas of manganese enrichment
are charted. The first of these is the tongue-shaped zone which extends northward
from the seaward end of Sarn Wallog. This coincides with the known distribution
of mixed coarse lithic fragments and sands and is delimited by the 400 ppm contour
The second is also outlined by the 400 ppm contour and constitutes a major offshore
enrichment zone extending from the cliffs near Borth northward to the vicinity of
Sarn Bwch. This area, however, contains some of the finest sands (Text-fig. 3) as
well as some of the best sorted ones (Text-fig. 6).

In the northern part, an elongate area of fine manganese-enriched sands occurs.

88 SEDIMENTATION IN NORTHERN CARDIGAN BAY

(oR a Sas ee sh 7/

KILOMETRES

Mawddach
Estuary. P

BA — BADRIG
BW — BWCH
WA — WALLOG

Fic. 28. Distribution of manganese based on spectrographic data. Contour interval
100 ppm.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 89

With one exception, all of the samples within a 5 km. radius of the mouth of the
Mawddach estuary at Barmouth are reported to have less than 500 ppm manganese.
Indeed, the littoral samples between Ty-wen and Eglwys Llanaber are noticeably
deficient in manganese as none contains more than 300 ppm.

The data do not show that a systematic distribution of manganese exists within
the Mawddach estuary. On the other hand, sediments in the Dyfi estuary are
enriched both along its south shore as well as in the immediate vicinity of Aberdyfi.
River samples associated with streams draining into the estuaries are deficient in
manganese when compared with the marine sands.

Sodium. The distribution of sodium expressed as salients is limited to three areas :
first, a band of sediments extending north from the outer end of Sarn Wallog ;
second, a band extending northward from the vicinity of Borth ; and, third, a
small area of concentration off Towyn. For the rest of the northern part of Cardigan
Bay, there is no pronounced geographic pattern to the distribution of sodium, and
for most stations values between 0-30 and 0:50 % are reported.

Nickel. At only one station does nickel occur in an amount less than 10-0 ppm,
and although, for most of the bay, nickel is present in the sediments in amounts
between Io and 20 ppm, four areas of pronounced nickel enrichment are known
(Text-fig 29). First, an area delimited by the 20 ppm contour extending north-
ward from the end of Sarn Wallog is of local importance, and correlates with known
coarser grained sands. Second, a narrow zone extending northward and paralleling
the coast is developed between the cliffs south of Borth and the mouth of the Dyfi
estuary. Third, a local zone of enrichment occurs inshore between Towyn and,
crossing the sarn, the vicinity of the eroding cliffs north of Tal-y-garreg. Fourth,
a distributive zone extending south from the shoal end of Sarn Badrig has been
charted.

Lead. Four major zones of enrichment occur in the northern part of Cardigan
Bay (Text-fig. 30), which provide clues to the origin and distribution of bay sediments.
The most well defined area of lead accumulation is that which extends northwest-
ward from the cliff exposures near Borth. The contours bifurcate, suggesting that,
while the lead is related to a common source, it is distributed in two directions,
reflecting variations in the local tidal current regime. Correlating with increased
grain size and poorer sorting, a second zone of enrichment extends northward from
the seaward end of Sarn Wallog. Though the geochemical discussion of lead is
presented later, it may be pointed out here that this same area is also one with higher
amounts of phyllosilicates. Extending northward from Towyn and beyond Sarn
Bwch into the middle of the north part of the bay is another zone in which sediments
contain this trace element in amounts exceeding 20 ppm. The distribution of lead
in this area indicates that sands entering the bay near Towyn are transported north-
ward over the shoaling, landward end of Sarn Bwch. A zone of enrichment, probably
associated with tidal currents and transport from beyond Sarn Badrig, occurs in
the far northern part of Cardigan Bay.

Littoral sands on either side of the estuary mouth at Barmouth are noticeably
deficient in lead, as all samples collected close to the coast, except one, contained

go SEDIMENTATION IN NORTHERN CARDIGAN BAY

SAN SSG 7 BreS Oe
———
KILOMETRES

BA — BADRIG
BW —- BWCH
WA- WALLOG

Fic. 29. Distribution of nickel based on spectrographic data. Contour interval 10 ppm.

SEDIMENTATION IN NORTHERN CARDIGAN BAY QI

fe)

Piteueom 400s Ouse ee On Oty

KILOMETRES

C) \Abendyti :

Ne

BA - BADRIG
BW - BWCH
WA - WALLOG

Aberystwyth

Fic. 30. Distribution of lead based on spectrographic data. Contour interval 10 ppm.

92 SEDIMENTATION IN NORTHERN CARDIGAN BAY

less than 1oppm Pb. In the Dyfi estuary lead is present in amounts exceeding
40 ppm in the marsh deposits along the south side of the estuary. Sediment samples
collected from the nearby Afon Leri and Afon Clettwr were analysed and found to
contain lead in amounts between 118 ppm and 186 ppm, thus more than twice the
amount reported for the marsh deposits near their mouths.

In summary, the major areas of lead concentration are related to coastal erosion,
to coarse lithic grains eroded from Sarn Wallog and to an, as yet, unidentified source
near Sarn Badrig.

Scandium. Over most of the offshore survey, scandium occurs in amounts
between 4 and 10 ppm. Nevertheless, three areas where scandium exceeds 10 ppm
are important to understanding sediment sources. The first of these is located
adjacent to the coast in the vicinity of Borth. The second is the tongue-shaped
area extending northward from the seaward end of Sarn Wallog. Within this area,
only two stations (190 and 14) contain more than 10 ppm; however, several other
stations approach this value. Thirdly, northwest of Towyn, Stations 62, 33 and
34 are reported with values exceeding I0 ppm, and, as such, the area, although
small, is important. Although several samples from the estuaries have values in
excess of 10 ppm Sc, there is no evidence seaward of the estuary mouths that
scandium is transported in sands through the estuaries and into the bay.

Strontium. Strontium is enhanced in only two areas in the northern part of
Cardigan Bay. The first of these extends northward from the seaward end of
Sarn Wallog where some of the highest strontium values are reported. This same
general area is also one of sediments with abundant shell fragments. The second
area of distributive importance is that located in the far north of the survey adjacent
to Sarn Badrig. For most of the offshore sands, strontium is present in amounts
varying between 90and 150 ppm. Littoralsands are, however, deficient in strontium
as are the sediments in the upper estuaries. Samples from river beds associated
with both the Dyfi drainage area and the Mawddach network contain less strontium,
on the whole, than do the offshore sands.

Titanium. Sediments containing more than 2,000 ppm of titanium are essenti-
ally restricted to two major areas (Text-fig. 31). The first is an inshore province
approximately 8 km. wide which extends northward from Sarn Wallog to Sarn
Bwch ; it gradually narrows after crossing the sarn and terminates in the vicinity
of Ty-wen. This area is largely one of well sorted fine sands actively swept by
bottom currents. The second is that elongate zone extending from Sarn Badrig
in the north corner of the survey southward to near the end of Sarn Bwch. Within
this narrow area are some of the highest values reported. These areas, particularly
the northern one, are in regions where the sediments are deficient in phyllosilicates
and, thus, much of the titanium present is related to other mineralic hosts, namely
the heavy minerals. There is no pronounced distribution pattern for titanium in
the estuaries. However, data for the several river samples show that titanium is
present in stream deposits in excess of 4,000 ppm, reflecting, of course, the abundant

SEDIMENTATION IN NORTHERN CARDIGAN BAY 93

ORiia ae A Om Oni Bro Oni
ee it
KILOMETRES

* of OTy-wen

BA — BADRIG
BW — BWCH
WA — WALLOG

Fic. 31. Distribution of titanium based on spectrographic data. Contour interval
I,000 ppm. (values x 10%).

04 SEDIMENTATION IN NORTHERN CARDIGAN BAY

(Oo) fl ee ey GAG Toei
_—— oo |
KILOMETRES

\ Barmouth ¢/°

“SS 7 et (a

BA — BADRIG
BW- BWCH
WA —- WALLOG

Fic. 32. Distribution of zirconium based on spectrographic data. Restricted contours
of 500 and 1,000 p.p.m.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 95

lithic fragments. In all, titanium is enhanced in current swept, well sorted, fine
grained sands.

Vanadium. In seeking a reasonable concentration level for describing the distri-
bution of vanadium, values of 60 ppm or higher are considered significant. On
this basis three areas are delimited, namely : a zone extending north from the
seaward end of Sarn Wallog ; an irregular zone extending northwestward from Borth,
and finally a zone in the northern area of the survey adjacent to Sarn Badrig. Several
samples contained more than 60 ppm vanadium, in particular those in the vicinity
of Sarn Bwch, but none are believed to be significant for establishing distributive
patterns. Likewise, data for the estuary samples do not suggest any systematic
distribution pattern within these smaller bodies of water.

Zirconium. The charted distribution of zirconium (Text-fig. 32) shows that while
this trace element is present at all stations, it varies within extremely wide limits.
From a low of 51 ppm to a high of 10,100 ppm, it would first seem that the spread
of data is too great to be used in establishing any logical distribution system. None
of the minerals determined by X-ray diffraction shows such a proportionately
wide range ; nevertheless, by selecting the 500 p.p.m. contour as a minimum line
of distributive significance, it can be shown that zirconium is enriched in two areas.
The first of these is an irregularly shaped area which extends northwestward from
Borth and is related to fine, well sorted sand. The second area is located in the
northern part of the survey, bounded by the inshore part of Sarn Badrig and the
coast between Barmouth, and the terminal end of this same sarn. The area forms,
in outline, an inverted U-shaped zone.

Data for the several estuary stations do not establish any local distributive pat-
terns. Since the data for the several river samples do not exceed 255 ppm, it may
be assumed that, on the whole, offshore sediments are much enriched in this trace
element in comparison with the river and estuary deposits; that zirconium is high
in these bay deposits in general, and that zirconium enrichment is indicative of high
energy sedimentation.

VELTs DISCUSSION

In the previous pages of this report Cardigan Bay sediments have been described
in terms of their textural, petrographic, mineralogical and chemical characteristics,
and individual dispersal and distribution patterns have been deduced from these
data. Attention must now be given to significant inter-parameter relationships,
and the environment of sedimentation. Composite dispersal and distribution charts
show the general sedimentation pattern, as well as the significant factors which have
locally influenced it.

In this study some comparisons are necessary, and consequently data for other
Recent sediments, as well as some Welsh reference rocks (Tables I and II) have
been included. No attempt has been made to compare trace-element and mineral
data for which there is no textural information, or equivalent reference of comparison.
It is, for example, meaningless to seek similarities of element abundance between

96 SEDIMENTATION IN NORTHERN CARDIGAN BAY

coarse grained shallow water sands and abyssal clays. The following topics are
selected examples and many further studies may be made from the reported data.

(a) Sediment Dispersal

The overall dispersal of sediments in northern Cardigan Bay is clearly shown by
a composite dispersal chart (Text-fig. 33). The most obvious dispersal trend is
that extending northward from the cliffs at Borth ; this suggests that numerous
grain types are transported northward from this local source. Furthermore, trans-
port continues northward past the mouth of the Dyfi estuary and along the coast
west of Aberdyfi. Of importance is the fact that several dispersal/transport vectors
charted for the Dyfi estuary mouth suggest that several grain types enter the estuary
from the bay. Furthermore, Dyfi estuary data suggest that some fine sand is
transported up the estuary and that some is deposited along the south shore of the
estuary. Locally, minor transport zones totally within the estuary may exist, and
the erosion of exposed rocks along the north shore could provide a source for the
small amount of grains apparently introduced there. The significant observation
is that there is no great dispersal from the estuary into the bay.

For the area north of Aberdyfi, in the vicinity of Towyn, the network of vectors
suggests that the eroding cliffs at Towyn are important local sources of detritus.
These vectors, chiefly oriented north or northwest, show that the dispersal of intro-
duced detritus is in that general direction. Furthermore, at least four vectors show
that sand is being actively transported over the shoaling end of Sarn Bwch. The
proximity of vector terminations in the vicinity of the sarn suggests that this same
situation may well occur for several other lithic and quartz grain types. Some sand
is transported in a westerly direction, and may in part be passed around the end of
Sarn Bwch and in part be transported seaward.

North of Sarn Bwch, dispersal vectors reveal that at least ten types of grains are
dispersed from the vicinity of the eroding cliffs along that part of the coast (Text-fig.
33). In this general vicinity, but seaward some 3 km., a trend of northward trans-
port is apparent. This continues uninterrupted northward along the coast, passing
offshore from Barmouth and the mouth of the Mawddach estuary. Only four vectors
are charted for sediments within the Mawddach estuary itself. Three of these
represent sands transported into the estuary from the bay, and one represents
dispersal from the estuary seaward. It is important that there is no obvious
alteration of the prevailing northward transport pattern immediately offshore
from Barmouth. If any significant amount of detritus were being delivered sea-
ward, it would have been charted and its presence recognized. Northward beyond
Barmouth, the composite dispersal parallels the coastline; however, in approaching
the shoaling Sarn Badrig region, dispersal vectors show that, in part, transport is
to the west. In this same region, several vectors suggest that some sand is entering
from beyond Sarn Badrig and is being introduced into the established distribution
system, even while a minor amount of sand is dispersed northward and close to the
shore. That this is the case is further substantiated by the known accumulation
of sand along the coast between Eglwys Llanaber and Mochras Point.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 97

OMe. Sra) 1G 72 Soon s
KILOMETRES

Mawddach | ::
Estuary os) ei

Barmouth 5 af

BA —- BADRIG
BW - BWCH
WA- WALLOG

Fic. 33. Chart showing dispersal vectors for each of the petrographically determined
grain types.

98 SEDIMENTATION IN NORTHERN CARDIGAN BAY

A major zone of sand dispersal is established in the middle of the bay (Text-fig.
33). This narrow area is considered one of the most active transport zones within
the survey. Itis, furthermore, related to the shoaling bathymetric salient (Text-
fig. 1) for this region, which, in turn, reflects the net accumulation of detritus there.

For the western margin of the present survey, dispersal vectors show a diminution
eastward, 1.e., dispersal from a source seaward. However, firm conclusions regarding
this far western overlap zone must await subsequent studies beyond the charted
limits of the present survey, and it must be remembered that much of the area west
of the sarns is composed of somewhat coarser clastics (Text-fig. 4). Eastward
diminution in this region of coarser sediments could be a reflection of simple mixing
of coarse detritus with the abundant fine sands found closer inshore ; the data are
not conclusive.

In the southwest corner of the region, six vectors all very nearly parallel, are
evidence that some sand is introduced into the area from the south, and beyond
the seaward terminus of Sarn Wallog. Also, in the vicinity of Sarn Wallog, numerous
vectors with “ greater-than-average’’’ magnitude are charted for a narrow zone
extending northward from the end of this sarn. Since several of these are for lithic
grains, it appears most reasonable to explain this zone as one of dispersal and trans-
port away from the outer eroding terminus of the sarn. An alternative explanation,
which may be substantiated only by study of deposits south of this survey, is that
some of these grains are transported around the end of the sarn from a source, or
sources, south of Aberystwyth. That some sand is transported over Sarn Wallog
from the south is evident from the vectors west of Aberystwyth.

A number of east/west axial trend lines occur in the middle part of the chart.
Although several of these may be interpreted as westerly-pointing vectors they are
designated simply as zones of transport.

In summary, the composite dispersal plan for the grain types (Text-fig. 33) reveals
that the sands in northern Cardigan Bay are current transported in a definite and
orderly pattern. Briefly, their movement is northward along the coast to near
Sarn Badrig, thence south/southwest to the vicinity of the middle part of the area.
Sands are also introduced from beyond Sarn Badrig and from beyond the terminus
of Sarn Wallog. Moreover, the net transport of sand is into the estuaries and not
out of them. The rivers contribute virtually nothing to the bay deposits.

(b) Distribution of chemical elements

By preparing a composite vector distribution chart (Text-fig. 34), the overall
plan of element distribution is clearly shown ; it provides evidence of environmental
and local source influences. In the absence of adequate bottom current data, the
composite vector chart of chemical elements provides evidence of the effective relative
magnitude of tidal currents and their direction. Obviously, chemical elements by
themselves are not dispersed, but are distributed according to the dispersal of their
mineralic hosts. Nevertheless, the common geochemical-mineralogical associations
of most elements, including the trace elements, are fairly well understood (Rankama

SEDIMENTATION IN NORTHERN CARDIGAN BAY 99

& Sahama, 1950 ; Turekian & Wedepohl, 1961), and it is possible to use charted
chemical data for interpreting the major distribution scheme.

As in the previous case for grain dispersal, element distribution vectors (Text-fig.
34) charted for the region just west of Borth show pronounced northward alignment.
Furthermore, the vectors are largely confined, not only west of Borth, but for most
of the survey, within narrow zones, much more so than are grain dispersal vectors.
Vectors for the area immediately west of Towyn confirm a distribution centre and
northwestward transport from there. Element vectors provide additional evidence
that sediments are transported across the shoaling end of Sarn Bwch and thence
northward along the coast west of Ty-wen. Vectors charted for the region of sands
adjacent to the cliffs north of Sarn Bwch (PI. 2, fig. 2) show that sediments originate
there and are transported northward. Within the northern part of this survey, a
pronounced transport salient parallels the coast off Barmouth and turns westward
on approaching the shoaling end of Sarn Badrig, whereupon the transport zone is
directed in a south/southwesterly direction and terminates near the end of Sarn
Bwch. This salient correlates with the charted lobe of the 10 fathom bathymetric
contour (Text-fig. 2).

Text-fig. 34 shows two clusters of distribution vectors adjacent to Sarn Badrig.
Those in the vicinity of the shoaling end of Sarn Badrig are directed in a southward
direction and join the mid-day salient. For deeper water, other vectors suggest the
introduction of sands from beyond Sarn Badrig in that vicinity. Numerous vectors,
pointing eastward, suggest that some detritus is brought into the area from beyond
the western margin of the survey. A suite of vectors extending northwestward
and northward from the outer reaches of Sarn Wallog defines a major mid-bay
zone of sediment transport.

In the south/central part of the survey, several lines designating zones of distribu-
tion may be interpreted as transport either north or south and either east or west;
it is not possible to define a preferred direction. These mid-bay “ trend lines ”’
are, however, in alignment with a known gap in Sarn Wallog. Distribution of
chemical elements within and at the mouth of each of the estuaries supports the
previous suggestion that estuary infilling is by sediment from the bay. No chemical
vectors could be plotted for distribution seaward from the Dyfi estuary, and only
one vector is charted for distribution seaward at the mouth of the Mawddach
estuary.

The current-influenced accessory mineral enrichment results in proportionately
higher values for several trace elements and shows that preferential enrichment
takes place in the paths of strong tidal currents. Zirconium enrichment, for example,
occurs in the fine sands, and since there is no concomitant increase for any gross
mineralogical component, distribution vectors for those elements normally associated
with accessories are hydraulically related (cf., Rittenhouse, 1943). Thus, while the
composite element distribution plan, in part, reflects tidal current/mineralogical
associations, it does not always signify true dispersal. In Cardigan Bay, grain
dispersal and element distribution plans are obviously related, however. While the
composite grain dispersal plan (Text-fig. 33) is undoubtedly the better overall
indicator of sediment source and transport within the region, the element distribu-

MINER. 2, 2. 7)

FIG. 34.

SEDIMENTATION IN NORTHERN CARDIGAN BAY

(0) Se sre 7A} 1
_—— SS Se

KILOMETRES

Mawddach +
Estuary). 4
Barmouth” fei.
ee,

BA — BADRIG
BW - BWCH
WA - WALLOG

Chart showing distribution vectors for the spectrographically determined chemical
elements.

SEDIMENTATION IN NORTHERN CARDIGAN BAY IOI

tion plan (Text-fig. 34) is the better indicator of direction and relative magnitude of
tidal currents, although many elements are dispersal controlled.

(c) Compositional classification of sediments

Two schemes (van Andel, 1958 ; Pettijohn, 1949) were used to establish the
compositional classification of the Cardigan Bay deposits (Text-figs. 35, 36).

Recent reviewers (Huckenholz, 1963 ; Klein, 1963) of sandstone classifications
emphasize that most schemes are based on petrographically determined end-members
and that one pole is generally assigned to lithic fragments, thus limiting the appli-
cability of diffraction data, particularly where lithic fragments are abundant. On
the other hand, a classification based on gross mineralogy is useful for comparing
chemical data, since the distribution of elements related to gross mineralogy is
meaningful only when so compared, regardless of the grain types present. The
better sorted a deposit is and the more its constituent minerals are found as single
grains, then the more worthwhile the gross (diffraction based) classification and
the more comparable it is with a petrographically based classification. In a study
concerned with establishing the petrographic nature of deposits, a genetically
oriented classification (Van Andel, 1958) would suffice. However, sedimentologists
need a classification based upon more rigid, quantitative, compositional limits and
less subject to inherent gross variation resulting from specific or interpretive varia-
tions of the constituent grains themselves. If, for example, several hundred indi-
vidual grains, all of which contained-a small fragment of a phyllosilicate mineral,
were counted, the sediment would be classified as a “ greywacke’”’ according to
Van Andel’s scheme. It is clear that the chemical and X-ray diffraction data
for this sediment could not be compared with data for clay-rich greywackes. Thus,
some reasonable approach must be taken in classifying unconsolidated detritus lest
the resulting nomenclature be wholly unrepresentative in fact as well as in concept.
Furthermore, as the use of chemical and X-ray diffraction data becomes more
commonplace each year, it is imperative that a scheme based on the gross mineralogy,
regardless of its petrographic distribution, should be considered. This is particu-
larly true in comparing chemical data for the several types of deposits, for in many
instances it is not feasible to subject samples to partition analysis in order to establish
component limits and geochemical/mineralogical associations. Indeed, in comparing
Recent with ancient sediments, gross diffraction analysis and gross spectrochemical
analysis may well provide the most useful data for approaching comparative sedi-
mentological investigations.

Text-fig. 35 shows the Cardigan Bay sediments plotted according to the van
Andel (petrographic) scheme in which clay is not an end-member in spite of its geo-
chemical and environmental importance. According to this scheme, most of the
Cardigan Bay sands fall into the sub-greywacke category, with a few assigned to
the greywacke category. Unfortunately, this terminology is misleading because
the greywackes were derived from the Aberystwyth grits in the first place. If
geochemical correlations based on the van Andel scheme were made, it would be
difficult to assign proportionate values to those elements associated with matrix
clays, or with combined lithic/mica grains, regardless of their textural state. The

102 SEDIMENTATION IN NORTHERN CARDIGAN BAY

QUARTZ

FELDSPARS 50 ROCK FRAGS

Fic. 35. Cardigan Bay sediments plotted according to the van Andel (1958) classification
scheme. Note that most bay sands are classified as sub-greywackes. This scheme is
based on petrographic data. Types are as follows: Q, quartzose sand; SG, sub-grey-
wacke ; G, greywacke; SA, sub-arkose; and A, arkose.

van Andel scheme does show, however, that relative to the individual feldspar
grains, the effective variation is simply a changing ratio between quartz and lithic
fragments.

Text-fig. 36, based on the Pettijohn system, is plotted using the end-members
quartz, feldspar and clay, regardless of their texture or petrographic combinations.
The reported diffraction data were used to plot each sample in this scheme, and
here we see a different categorization for the majority of sediments in northern
Cardigan Bay. Unlike the van Andel scheme, whereby most of the deposits are
classified as sub-greywackes, the Pettijohn category in which they fall is that of
quartzose sands with some few samples, about ten, classified as feldspathic sands.
The interesting feature about this classification is that clay (phyllosilicates) is
important, although it is completely within the lithic fragments. Moreover, this

SEDIMENTATION IN NORTHERN CARDIGAN BAY 103
CLAYS

FELDSPARS 25 10 QUARTZ

Fic. 36. Cardigan Bay sediments classified according to the basic Pettijohn (1957) scheme.
Samples plotted here are based on X-ray data for end-member assignment (cf. Moore,
1963). Note that in this system most of the bay deposits are classified as quartzose sands.

scheme is ideally suited for comparing trace element data and other mineralogically
related parameters, particularly when sediment terminology alone is reported, and
it has been used for classifying other Recent, shallow water sediments (Moore, 1963).

A descriptive petrographic term, not necessarily an entire classification scheme,
which may be used to modify the textural description, is useful, particularly for
providing more complete descriptions and for signifying the abundance of lithic
grains relative to quartz grains. In keeping with this need, a compositional modifier
based on petrographic determinations is suggested. This is based on the ratio of
quartz grains to lithic grains. A ratio of 9-0 or greater signifies quartzose ; between
g-0 and 3:0, sub-lithic ; less than 3-0, lithic. In the absence of appreciable quantities
of detrital feldspar and clay, this modifier suffices to define the general petrographic/
compositional character of the deposit. Used in conjunction with the textural and
sorting terms, it provides a descriptive terminology readily understood by others.

104 SEDIMENTATION IN NORTHERN CARDIGAN BAY

In short, the van Andel scheme is recommended for any comparison based on,
and limited to, petrographic variables, whereas the Pettijohn scheme is recom-
mended for comparisons relating gross mineralogical and geochemical associations,
needing a common reference of classification. By the former, most bay sands are
termed “‘ sub-greywackes’”’ ; by the latter, “‘ quartzose sands ”’.

(d) Textural and mineralogical relationships

Text-fig. 37 shows the correlation of grain size (6 Mz) with total phyllosilicates,
i.e., mica and chlorite, as determined by X-ray analysis. For those samples coarser
than 2 ¢ Mz, there is a corresponding increase in the amount of total phyllosilicates
ascribed to increasing amounts of lithic fragments contained in the coarser grained
deposits. Samples containing small amounts of phyllosilicates, i.e., less than 10%,
are generally restricted to the size range between I and 3 ¢ Mz. The lowest phyllo-
silicate contents are found in sediments between about 2:0 and 2:5¢ Mz. One
might deduce from this that lithic grains, in general, are removed from the fine sands
by the abrasive action of transport by bottom currents, inasmuch as these same fine
sands are those highest in quartz.

Folk (1961, p. 125) suggests that the mineralogy of sediments may be controlled
by grain size through preferential sorting or by the composition at the source. In
the present study we have found that there is an overlap of these two causal effects.
In the first instance, most of the lithic fragments are similar in composition to the
adjacent rocks of Wales, and thus disaggregation into various sizes does little more
than effect a clastic state. Study of the data, as well as careful inspection of thin-
sections, has failed to show that there is any appreciable diagenesis of the lithic
fragments encountered in these bay deposits. Furthermore, little opportunity
exists for deposition of fine mica and chlorite flakes after their separation by mechani-
cal degradation of larger lithic clastics. The environmental energy regime for this
area of Cardigan Bay is such that strong currents remove all sizes below the very
fine grained sands and their hydraulic equivalents. It should be noted too that the
diagenesis of mica and chlorite is subject to the same laws of physical chemistry
as any other particulate matter, i.e., a large specific surface area is necessary to
enhance the rate as well as the completeness of chemical alteration. The short
length of geologic time that these phyllosilicate-bearing lithic grains have been
submerged in Cardigan Bay waters and their inherently large grain size and small
surface area eliminates the possibility of advanced diagenetic (chemical) alteration.

Nowhere in the reported data do values for the fine fractions suggest that phyllo-
silicates occur as clay size, individual mineral grains ; thus, all phyllosilicates are
functionally, as well as genetically, related to lithic fragments. Indeed, for those
samples having values for coarse sizes exceeding 1 ¢ Mz, the total phyllosilicates are
always indicative of large fragments of previously existing rocks, primarily slates
and arenites. Further Text-fig. 37 suggests that, for decreasing grain size below
fine sands (higher positive ¢ Mz values), there is only a slight increase in total
phyllosilicates.

Careful petrographic study of the various samples shows that the higher phyllo-

SEDIMENTATION IN NORTHERN CARDIGAN BAY 105

(X-RAY)

°lo

Y)
WW
=
fol
=
<
VD
x
zi
ZL

PHYLLOSILICATES

Fic. 37. Relationship between texture (¢ Mz) and total phyllosilicates (1o A muscovite
and 14 A chlorite). Note the correlation between increasing grain size and increasing
mica-chlorite content, the latter being related to lithic fragments.

silicate values at marsh stations in the estuaries are related to accumulations of very
small slate fragments and not to clay detritus. In fact, it is only at these few isolated
estuary sites that apparent enrichment of these two clay varieties (ro A and 14 A)

106 SEDIMENTATION IN NORTHERN CARDIGAN BAY

may be realistically interpreted as true deposition. The hydraulic entrapment is
caused by protection of the site by Spartina or other marsh grass, and by locally
weak tidal currents. That these estuary sites are local areas of low-energy deposition
is further substantiated by the geochemical data, for instance, zirconium is not
enriched there. In summary, phyllosilicates correlate with increasingly negative
@ Mz values and thus with coarser texture, and only rarely with decreasing grain sizes
smaller than the 3 ¢ Mz boundary.

Folk (1961) has pointed out that feldspar in reasonably well sorted marine sands
is most abundant in the size ranges between about 2 and3 ¢. This textural/mineralo-
gical relationship is also found for the Cardigan Bay sands. Inasmuch as the thin
section studies show that feldspars predominantly occur as individual grains and
are usually smaller than the host fraction, the use of diffraction data is instructive.
A plot of ¢ Mz with total feldspar in sediments smaller than 1-5 ¢ shows an increase
in feldspar with decreasing grain size. For the fine sands as a group the minima
for feldspar (approximately 3°) occur near the 2¢ Mz boundary (Text-fig. 38),
while the maxima (in excess of 10%) are noted for those samples near the 3 ¢ Mz
boundary. In considering those sediments coarser than 1-5 ¢ Mz, there appears to
be a linear increase at least through the very coarse sand ranges. Since this observa-
tion is based only on about twenty samples, it is questionable whether any definite
conclusions can be reached. Moreover, feldspar in the coarser grained deposits is
related, in part, at least, to feldspar within rock fragments, and small grains of
feldspar result from abrasion within the environment of sedimentation and from
the grain sizes which prevail at the source.

Although quartz is considered in detail elsewhere in this report, it is useful from
an interpretive standpoint to consider the correlation of texture with total quartz
(Text-fig. 38). This plot shows that the data maxima are largely restricted to those
samples between 1-8 and 2:5 ¢ Mz, with the majority falling in the 2 to 2-5 ¢ Mzrange.
It is noteworthy that the coarser range of the fine sand interval is also the range
for phyllosilicate minima. This means that the most mineralogically mature sands
are those in the 2 to 2:5 ¢ range of the fine sand interval. With increasing grain
size, there is a decrease in the amount of quartz present, i.e., for those samples
exceeding —1¢ Mz, less than about 60% quartz is reported. Such a relationship
is logical in that these same coarse samples are those containing the greatest amount
of lithic fragments. Furthermore, for those samples with reported values between
2-5 and 4 ¢ Mz, thereisalsoa decrease in total quartz, which will be connected with an
increase in feldspars and phyllosilicates. The maximum sizes of many grains are,
however, controlled by their sizes in the original provenance suite. Though hydro-
thermal quartz veins present in the adjacent land mass could provide large quartz
fragments, the data do not show that they contribute much to the bay detritus.

The several quartz maxima as determined by X-ray diffraction confirm petro-
graphic observations, i.e., the most mineralogically mature sands are those between
20 and 2:5 ¢ Mz, i.e., largely the coarser range within the fine sand interval. While
X-ray data for total quartz are useful as indicators of gross mineralogical maturity
of the sediments, detailed study of quartz by petrographic methods is necessary in
order to establish its nature and dispersal.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 107

80

AMOUNT IN SAMPLE % (X-RAY)

erage
one cee e
oe —-Hitftee °

Fic. 38. Relationship between texture (¢ Mz) and total quartz (upper plot), and between

texture (¢ Mz) and total feldspar content (lower plot).

Note the definite correlation

between decreasing grain size and increasing feldspar for the fine sands (2-3 ¢ Mz), and
the quartz maxima associated with medium and fine sands.

108 SEDIMENTATION IN NORTHERN CARDIGAN BAY

(e) Textural and petrographical relationships

Inasmuch as considerable emphasis is placed on the petrographical character of
the several quartz types and their dispersal within the region, it was necessary to
test statistically the relationship between the several quartz varieties and the texture
of the constituent deposits. In no instance was an obvious correlation between
quartz type and grain size found. The importance of “‘ randomness” between
specific quartz varieties and texture cannot be over emphasized, since for the dis-
persal plans to be significant, there must be no correlation between quartz type and
grain size. On the other hand, a correlation does exist between texture and lithic
fragments for those samples with median diameters exceeding about 1 mm. As
these deposits constitute a relatively small portion of the bay floor cover, this
correlation is seemingly important. However, within the well sorted fine grained
sand, a single Wentworth grade, no correlation exists between total lithic fragments
and intra-grade textural variations.

During the course of grain counting, it was noticed that most of the orthoclase
occurred as sub-rounded grains which were usually smaller in size than the host
(quartz) fraction. In most samples, plagioclase grains followed this same pattern.
This petrographic observation agrees with that made previously from X-ray and
textural data.

(f) Relationships between chemistry and texture

For some elements, there are correlations between the texture of the Cardigan
Bay sands and element abundance. Two examples of this are the correlation of
zirconium and median diameter, and the correlation of titanium and median diameter.
Text-fig. 39, a plot of zirconium and median diameter data, shows that zirconium
enrichment is restricted entirely to the fine sands. Since the reported mineralogical
data do not show proportionate phyllosilicate increases (the only other possible
carrier of zirconium), the enrichment of this element results from the presence of
the accessory mineral zircon. Zircon is abundant in Cardigan Bay sands.

Zirconium values generally increase with decreasing grain size, from low values
of about 150 ppm for those samples with a median diameter exceeding 0-7 mm.,
to high values of over 3,000 ppm for certain fine sands with Md values of slightly less
than 0-2 mm. (Text-fig. 39). Indeed, the rate of increase of zirconium with decreas-
ing grain size is greatest in this fine sand range. This relationship is reasonable in
light of previous research by Rittenhouse (1943), who found that zircon, because of
its high specific gravity, could be present as finer grained clastics within somewhat
coarser grained deposits. He found that the Md for zircon grains was about 0-07
mm. for a host fraction with an Md of about 0-18 to 0:20 mm. The present inter-
pretation of Cardigan Bay sediments is that the zircon grains, presumably weathered
from such sources as the nearby grits, are size equivalent to the host sand and,
furthermore, that the differential in transport is such that the quartz grains are
transported at a slightly higher rate than are zircon grains. This mechanism, which
admittedly may not always be balanced, provides for net enrichment of zircon
grains, 1.e., the zircons accumulate as a minor lag concentrate due mainly to preferen-

SEDIMENTATION IN NORTHERN CARDIGAN BAY 109

T Cais hh

ZIRCONIUM ppm

¢ CARDIGAN BAY DEPOSITS
O RIVER DEPOSITS
4 BUZZARDS BAY

wW
(2)
1e)
(e)

Fic. 39. Relationship between the trace element zirconium and texture (Md). Here
the Cardigan Bay deposits are seen to have pronounced enrichment of Zirconium in com-
parison with certain other similar textured sands. Zirconium enrichment in Cardigan
Bay sands is ascribed to local concentrations of zircon.

tial removal of some host constituents. Support for the foregoing interpretation is
found in the work of Inman (1949), who observed that fine sand is the most easily
moved of the several sand grades, and that the coarser and finer grades, being
moved by surface creep, tend to lag behind. Had the enrichment of zirconium been
reported for only one or two samples, the data could very well be regarded as ano-
malous, but with so many samples in the fine sand range with zirconium exceeding
500 ppm, they must be considered true concentrates, sedimentologically, if not
economically. Furthermore, the distribution of zirconium (Text-fig. 32) shows that
this element is concentrated in the paths of established grain dispersal and element
distribution zones (Text-figs. 33, 34), confirming the environmental tidal energy
control.

In order to compare these zirconium-rich deposits with others from a shallow
water environment but of known active deposition, the author has selected data
for certain other Recent sand samples of equivalent texture (Moore, 1963). These
data for a low-energy environment on the Atlantic coast of America also show an
increase of zirconium with decreasing grain size (Text-fig. 39), with the highest

110 SEDIMENTATION IN NORTHERN CARDIGAN BAY

zirconium values being reported for sands of approximately 0-15 mm. Md. While
sands in both areas of comparison are texturally alike, the Cardigan Bay (high
energy) sands are greatly enriched in zirconium over those sands deposited in a
normal (low energy) environment.

(Md) MEDIAN DIAMETER mm
1.0 3-0

Fic. 40. Plot of titanium with median diameter for Cardigan Bay deposits. Note the
increase in Ti with corresponding decrease in diameter for those sediments below 0-30 mm.
Md. Ti appears to be related to one or more of the accessory minerals in the fine sands,
and to lithic fragments in the coarse sands.

The foregoing interpretation is not invalidated by the work of Hirst (1962), who
concluded that zirconium distribution in sediments from the Gulf of Paria is probably
related to provenance rather than sorting of the sediments during transport. Gulf
of Paria deposits are not similar, generally, to the Cardigan Bay sediments because
the former are appreciably higher in clays, and, in fact, Hirst suggests (op. cit.,
p- 1,182) that the trace elements and most minor elements are structurally combined
within the lattices of the several clay minerals present there. Nevertheless, it is
interesting to consider his Zr data; he reports an average of 436 ppm for delta
sands and 413 ppm for platform sands. Neither of these values suggests that
abundant zirconium is structurally combined in phyllosilicates because (op. cit.,
p. 1,180) an average of 169 ppm is reported for the clay deposits. Furthermore,

SEDIMENTATION IN NORTHERN CARDIGAN BAY III

* CARDIGAN BAY SEDIMENTS
50 + ROCKS WALES
© RIVER DEPOSITS
4 BUZZARDS BAY (FINE SANDS) 4
/
b at
“er
20 Vi ae
& ve
oS, %e + +
=) siSsere st:
10 4 rae
<q
O)

A
* TOTAL PHYLLOSILICATES + FELDSPARS %.
10 20 50

Fic. 41. Correlation of gallium with the total of phyllosilicates plus feldspars. Here the
Cardigan Bay deposits and nearby river deposits show a pronounced linear relation.
Comparative deposits from Buzzards Bay also show a linear relation, but deficient in
gallium ; variation in the ro A/14 A ratio is a possible reason for the shift. Gallium replaces
aluminium in both phyllosilicates and feldspars.

van Andel & Postma (1954), in the original description of the Gulf of Paria sediments,
report (pp. 182-185) that zircon is the dominant heavy mineral in all the essentially
arenaceous deposits ; in fact, no other accessory mineral is present in quantities
exceeding that of zircon.

The present study suggests that spectrochemical data may provide a useful
indicator of the environment. The writer suggests that for normal, low-energy
deposition within a basin or on an open shelf, the rate of increase of zirconium with
the concomitant decrease in grain size is a gradual one, and that the range of average
values for zirconium in well sorted sands is probably of the order of 200 to 500 ppm.
This agrees with Pettijohn (1963, p. 11), who gives an approximate average range
of 200-250 ppm. However, in the case of mobile sands, a very sharp rate of increase
in the zirconium content with decrease in grain size below about 0-30 mm. Md is
significant evidence of preferential sorting by high energy tidal currents and, as
such, implies considerable transport. This interpretation does not rule out net

Te SEDIMENTATION IN NORTHERN CARDIGAN BAY

deposition for mobile, high-zirconium sands, but rather it provides a sound basis
for recognizing their relationship to a high-energy environment.

A similar relationship exists between titanium and median diameter (Text-fig. 40).
In this plot titanium increases both with decreasing grain size below about 0-3 mm.
Md and with increasing grain size above about this median diameter. The correla-
tion between increasing titanium and decreasing Md is a readily understood one in
that titanium is geochemically associated with two common accessory minerals,
rutile and ilmenite, insand. However, the increase of titanium with increasing grain
size demands a different interpretation. It is possible that some titanium is combined
within the lattices of phyllosilicate minerals which are known to be constituents of
the lithic fragments in these coarser grained sediments. In the present discussion,
the main interest is in the inverse regression between titanium and texture for the
fine sands in Cardigan Bay. Whether titanium is geochemically associated with
ilmenite or rutile, or both in these sands, is not an important point in itself, but rather
that it is associated with the heavy mineral suite. Thus, the reported titanium
values for Cardigan Bay sediments confirm the zirconium/Md relationship, i.e., some
elements are enriched in the fine sands and their mineralogical hosts are the heavy
minerals. Moreover, the relationships are sensitive to the physical energy environ-
ment, and thus may be used to interpret ancient sediments.

(g) Trace elements and gross mineralogy

Some trace elements in the Cardigan Bay deposits are related to gross mineralogical
constituents. An example of this relationship is expressed in Text-fig. 41, a correla-
tion plot of the element gallium with the total of combined phyllosilicates and
feldspars. Although these two mineral groups are in the minority for the bay
sediments, they are geochemically important as hosts for several trace elements.
Gallium data show a positive, linear relationship to the sum of phyllosilicate and
feldspar data percentages. The several gallium minima, i.e., values below about
5 ppm, show a slight spread, but the linear relationship is observable. For gallium
values above about 5 ppm, the relationship is a pronounced one. Sediments with
gallium in amounts above Io ppm also contain numerous rock fragments, and of
course, their phyllosilicates and feldspars. In fact, the distribution of gallium
provides important evidence of affinity with the phyllosilicates and feldspars rather
than with the accessory minerals. For instance, Station 240 with 5-2 ppm gallium
shows a total phyllosilicate/feldspar value of 13%. On the other hand, the zir-
conium content for Station 240 exceeds 10,000 ppm, and titanium at Station 240
is reported in excess of 20,000 ppm. These two data suggest an abundance of
heavy minerals, yet gallium remains low. Data for the reference rocks collected
in central and north Wales show a like linear relationship for this element and
its mineralogical associations although several data indicate less gallium. The
stream deposits plot on the same trend as for the bay samples. This is a reflection
of provenance.

Also plotted are selected data for Recent sediments of similar mineralogy and
texture from Buzzards Bay on the Atlantic coast of North America. These data
likewise show a linear distribution, but for the same amount of phyllosilicates and

SEDIMENTATION IN NORTHERN CARDIGAN BAY ma

feldspars, contain less gallium than the Cardigan Bay deposits. Since phyllosilicates
and feldspars are minerals rich in aluminium, the correlation of gallium with phyllo-
silicates and feldspars is a reasonable one because of the diadochy between the
trivalent ions Ga+-+-+ and Al+-+-+ (Rankama & Sahama, 1950, p. 720).

Hirst (1962), in his geochemical study of the Gulf of Paria sediments, suggested
that gallium as well as chromium, vanadium, copper, lead and barium, and, to a
lesser extent, cobalt, nickel and boron were brought to the basin of deposition struc-
turally combined in the lattices of clay minerals. Certainly, in the case of Ga, his
findings provide valuable insight into gallium-sediment geochemistry. Similarly,
Borisenok & Saukov (1960) in their study of the geochemical cycle of gallium report
that it is chiefly governed by its relationship with aluminium.

(h) Element-to-element relationships

Plots of element pairs are instructive and provide additional means for characteriz-
ing groups of sediments, for establishing element associations in nature, and for
reaching a better understanding of environmental influences. Of the numerous
possible combinations, five examples have been selected : titanium/iron ; calcium/
strontium ; zirconium/titanium ; scandium/gallium ; and cobalt/nickel.

Titanium|Ivon. A plot of Ti and Fe,O, data (Text-fig. 42) shows that a good
linear relationship exists between these two elements. For the fine sands (2-3 ¢ Mz)
in these deposits the Fe,O, content varies between 1-5 and 3%, whilst Ti varies
between 1,000 and 5,000 ppm. In addition, the Ti/Fe,O, values show that for a
small increase in Fe,O, there is a relatively large increase in the amount of titanium
present. The titanium in these sands is doubtless associated with accessory ilmenite
and rutile. This sharply defined linear regression represents high energy sands
which are well sorted, relatively mobile, and with only a small amount of phyllo-
silicate minerals present. For comparison, sands of similar texture and mineralogy
from Buzzards Bay (a low energy environment) have been plotted. These sands
show somewhat less iron, but equivalently higher titanium values, and although
they plot in a linear pattern, it has a different slope. For further comparison, the
data for the reference rocks from Wales are also presented. For these, there is a less
definite linearity to their plot, but they do contain almost twice as much iron as
some of the fine sands in the bay and yet contain no more titanium. Whether the
additional iron in the lithic samples is due to sulphides, or to iron-rich 14 A phyllo-
silicates is not known, but the latter is suspected.

In short, the steep regression for Ti and Fe,O, characterizes high energy sands.

Calcium|Strontium. The second example of element-to-element relationships is
that between calcium and strontium ; a relationship between a minor element (Ca)
and a trace element (Sr) and one of environmental significance. It has been reported
by Turekian & Kulp (1956) that strontium is present within calcite and aragonite
shells. Strontium, of course, may be, and usually is, a constituent trace element
of fossils and of sedimentary rocks in general. However, bioclastic fragments
provide, by far, the major amount of calcium carbonate in the Cardigan Bay deposits,
even though small grains of dolomite, usually sub-rounded euhedra, do occur in

114 SEDIMENTATION IN NORTHERN CARDIGAN BAY

re}

(e)

oO
TITANIUM ppm

¢ CARDIGAN BAY FINE SANDS
+ ROCKS WALES
4 BUZZARDS BAY FINE SANDS

IRON (AS Fez 03% )

0-5 1:0 2:0
Ht

Fic. 42. Plot of titanium with iron based on spectrochemically derived data. Note the
steep regression slope for the Cardigan Bay fine sands (high energy environment) in
comparison with the reduced slope for similar textured sands from a low energy environ-
ment.

some of the fine grained sands. This minor mineral is reported by X-ray diffraction
analysis in quantities of about 1% or less. Neglecting this small contribution of
dolomite, shell fragments constitute the major source for reported calcium. Calcium
and strontium are two elements which are added to the bottom sediments largely
by bioclastic material at the site of deposition. It is, however, significant that
strontium, although present, shows very little variation in samples (mainly fresh-
water) with a calcium content 1-5%. In these samples it is likely that the strontium
content is related to the phyllosilicates.

In seeking a comparison with other fine grained sand deposited in a shallow
water marine environment, data for selected sediments of equivalent texture have
been plotted (Text-fig. 43). Although only nineteen in number, these data do show
positive, correlative increases of strontium with increases in calcium below 1:5 %,
as well as for two stations exceeding 2:-5°%. In these sediments (all marine), even
the lower calcium and strontium values are related and increase together.

In summary, strontium is increased only for those sediments from the Cardigan

SEDIMENTATION IN NORTHERN CARDIGAN BAY 115

E
=
Qa
=
=|
=
Zz
O
or
ke
Y)

¢ CARDIGAN BAY
+ ROCKS WALES

o RIVER DEPOSITS
Q& BUZZARDS BAY

CALCIUM %

1:0

Fic. 43. Relationship between calcium and strontium. Both elements are enriched in
Cardigan Bay sediments due to the presence of bioclastics. Note the steep regression
for Cardigan Bay deposits (high energy environment) in comparison with the relatively
reduced slope for fine grained sands from a low energy environment. Calcium and
strontium have similar ionic radii, 0-99 and 1-12 respectively. Petrographic evidence
further suggests that calcium and strontium are two elements largely added in the
environment of sedimentation.

Bay area which contain bioclastic debris, and as such is indicative of the marine
environment.

Zirconium|Titanium. The ratio between zirconium and titanium is another
example of an element pair with good correlation. Here again, the bay data are
representative of a high energy environment (Text-fig. 44). Both of these elements
are, for these shallow bay sands, related to accessory minerals (zircon, ilmenite and
rutile). Moreover, the spread of titanium data between 1,000 ppm and 5,000 ppm
is such that an alternative relationship with phyllosilicates is unlikely. In comparing
the Cardigan Bay data with those from Buzzards Bay, a low energy site, the latter
possess less zirconium for the same relative Ti concentration, and they plot with a
lesser gradient. In short, the Cardigan Bay data exhibit a sharp Ti/Zr regression
when plotted, which serves to characterize them as high energy sands. An inverse
relation exists for many low energy deposits.

MINER, 2, 2. 8

116 SEDIMENTATION IN NORTHERN CARDIGAN BAY

* CARDIGAN BAY DEPOSITS
+ ROCKS WALES

5000 A BUZZARDS BAY SANDS
© RIVER DEPOSITS

+2000 . .°
. e
E oe
a 2 .
Liooo > OF M9 Ocak oF
>| aS .
ii = = Ae
i O tamer F
O . .
oe 4
500 N : 2 ray +
e e ~
oe
ea i
ef Age B
* ieie Fon + Sr
e $ + ° [e}
200 Be ie aay
© Me ‘
een Se er e

TITANIUM ppm

200

Fic. 44. Correlation between zirconium and titanium for Cardigan Bay deposits and other
comparative samples. Note the steep, positive regression for the Recent sediments,
with proportionately more zirconium in the high energy deposits. River samples and
reference rocks, both lithic in composition, however, show an inverse relationship between
zirconium and titanium. It is suggested that titanium in these is related to more than
one host, probably phyllosilicates and accessories.

Scandium|Galliwum. An example of correlation between two trace elements
related primarily to alumino-silicates is that of scandium and gallium (Text-fig. 45).
It has already been shown (Text-fig. 41) that gallium is correlative with the total
amount of phyllosilicates and feldspars present in a deposit. Scandium is similarly
related. Regardless of grain size or sorting of the sediment, a positive relationship
exists between scandium and gallium, and ratios of these two elements are controlled
by inherited mineralogy of the clastic fragments and ultimately the source rocks.
Inasmuch as data for reference samples of Welsh rocks plot on the same linear trend,
one may assume that a genetic relationship exists between the rocks and the bay
sediments.

Similar textured marine sands from Buzzards Bay, Massachusetts, also plot with
a linear Sc/Ga correlation, but they contain less scandium than the Cardigan Bay
sands. Study of the data suggests that variations in provenance mineralogy,
specifically the 10 A/r4 A clay ratio is the explanation. High scandium values are

SEDIMENTATION IN NORTHERN CARDIGAN BAY 117

* CARDIGAN BAY SANDS
+ ROCKS WALES
4 BUZZARDS BAY

GALLIUM ppm

10 20

Fic. 45. An example of the correlation of a trace element pair for Cardigan Bay deposits,
reference rocks from Wales, and Buzzards Bay sediments. Both the Cardigan Bay sands
and the Welsh rocks plot on the same line. The linear, but steep, regression for sedi-
ments from Buzzards Bay is thought to reflect variations in the 10 A/14 Aor other mineral
ratios. Ina few Cardigan Bay samples, scandium may be related to accessory minerals,
but the majority of scandium and gallium are related to aluminium bearing silicates.

reported for three Cardigan Bay samples with gallium values less than 5 ppm, and
these may be due to the presence of allanite, a mineral known to be present in the
nearby Welsh rocks (Bromley, 1964).

Cobalt/Nickel. Cardigan Bay sediments have a Co/Ni ratio of about 0-38 (graphic
determination), and reference rocks from North Wales, reported as a part of this
study, have a Co/Ni ratio of about 0-42, which closely agrees with the Co/Ni ratio
of 0-40 for a suite of shales from the Harlech Dome area nearby (Mohr, 1959). Hirst
(1962, p. 1163) reports Co/Ni ratios for several groups of sediments from the Gulf
of Paria which are in agreement with those above, e.g., 0-34 for delta sands, 0-49
for platform sands, 0-39 for green muds, 0-42 for clays and 0-34 for delta clays.
Hirst (op. cit., p. 1163) also pointed out that the mean ratio, about 0-40, represents
the near maximum for lattice combined Co and Niin sediments. The present study
with ratios very near to 0-40 is in agreement with the findings of Hirst and others.

118 SEDIMENTATION IN NORTHERN CARDIGAN BAY

In fact, Co/Ni ratios between 0-35 and 0-45 are considered, generally, as good
indicators of marine sediments.

Provided that trace elements are not lost during metamorphism (Turekian &
Wedepohl, 1961, p. 177), element ratios may be useful for establishing the original
rock type, and perhaps environment, of meta-sediments. The common repetition
of certain element-to-element ratios, such as Co/Ni and others previously referred
to, enhances the geochemical approach to the study of metamorphic rocks and their
genesis, but such an approach requires a firm understanding of the distribution of
critical chemical elements in modern sediments.

(7) Some general geochemical relationships

Numerous unfigured scatter diagrams (Krumbein & Pettijohn, 1938, p. 199)
were plotted for this study using the reported data. These plots in conjunction with
the findings of Weber & Middleton (1961), MacPherson (1958), Moore (1963), and
Potter et al. (1963) provide a basis for certain of the following interpretations.

Boron. The chemical association of boron is such that it is found in two host
minerals, namely, tourmaline, which has been observed in thin sections (e.g., B.M.
1964, 99; 127; and 174), and muscovite. For the latter some boron replaces
potassium in the ro A mica lattice, but it is doubtful that it is here an indicator of
salinity as suggested by Walker & Price (1963). Since the phyllosilicates in Cardigan
Bay sands are not detrital clays, it is unlikely that detailed boron/potassium studies
would yield accurate indices of salinity. This assumes, of course, the use of the
reported data for the whole sample. It is doubtful that the lithic fragments have
reached chemical equilibrium with Cardigan Bay water. In their study of lithified
sediments, Weber & Middleton (1961, p. 249) report some values exceeding I00 ppm
for boron. In their case, however, detrital clays were present in the samples.
Pettijohn (1963, p. 12) has pointed out that boron is seldom determined for sand-
stones, but that a reasonable estimate would be 25-35 ppm for the “average ”’
sandstone, which roughly corresponds to the amount found in the present survey.
However, in dispersal zones (Text-fig. 25), boron values exceed 40 ppm, as the result
of lag tourmaline in a high energy zone. Regardless of whether boron travels
with an accessory mineral or with micas bound in lithic fragments, its preferential
distribution is an important clue to sediment dispersal and to zones of sediment
transport.

Barium. In Cardigan Bay sands, barium is related to the phyllosilicates, to
feldspars, and in some samples to accessory minerals. The small amount of feldspar
present does not provide a wide enough spread to establish conclusively how much
barium is in the feldspars. Weber & Middleton (1961, p. 249) and Moore (1963, p.
541) have reported, however, that barium is related to feldspar. Weber & Middleton
(op. cit., p. 249) also point out that some barium is present in carbonate mineral
fragments. In the present study, this does not appear to be a general trend, since
for several samples with large amounts of bioclastic debris and with relatively high
barium values, there are, also, abundant lithic fragments present.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 119

Cobalt. The geochemical associations of cobalt are complex ones and its distribu-
tion in nature has not yet been fully explained (Carr & Turekian, 1961). In Cardigan
Bay, cobalt distribution is related to both phyllosilicate and accessory minerals, but
in the absence of partition analysis, we cannot establish the proportionate amount
in each. However, it appears that “ background ”’ cobalt is related to the mag-
nesium-bearing minerals, primarily chlorite, but that certain anomalous cobalt
values, e.g., Station 240 with 135 ppm Co, are related solely to accessory mineral
enrichment. Similar findings have been made by Weber & Middleton (1961),
including a pyrite/cobalt relationship. It is unlikely that pyrite is present in Cardi-
gan Bay sands, except as trace amounts in lithic fragments. Nevertheless, it would
be difficult to differentiate cobalt/sulphide from cobalt/phyllosilicate relationships
in lithic fragments using the reported data.

An interesting aspect of cobalt is that it is reported not to change with variation
in the metamorphic grade of regionally metamorphosed sedimentary rocks (Shaw,
1954). If this is correct and if sedimentologists can obtain sufficient quantities of
cobalt data on Recent sediments, one may speculate that the original environment
of metamorphosed sediments might some day be determined with reasonable
accuracy. Similarly, the use of this element in combination with another of equal
geochemical “ constancy ”’, e.g., nickel, may provide a useful geochemical tool for
stratigraphic correlation.

Chromium. Reference to Text-fig. 26 shows that chromium, like zirconium, is
distributed in zones where bottom currents are most active and that chromium is
largely enriched by accessory mineral concentration. A similar finding has been
made by Weber & Middleton (1961, p. 250) who consider that ilmenite and magnetite
contain most of this element. They do, however, suggest that clays carry some
chromium. While the Cardigan Bay samples are relatively depleted in phyllo-
silicates, the data suggest that a small amount of chromium is carried in that fraction.
Moore (1963) found that the amount of chromium in argillaceous sediments was a
good indicator of clay content, and while this observation is valid for low energy
deposits, it does not provide a similar index for well sorted high energy sands.
Chromium, in the latter, is concentrated in the accessory fraction and more directly
reflects prominent zones of sediment transport. Hirst (1962, pp. 1158-1161) in
his study of Gulf of Paria sediments found that Cr/Al and Cr/Fe ratios were moder-
ately constant throughout the basin, and suggested that chromium is associated
primarily with illite. For low energy argillaceous deposits like those in the Gulf of
Paria (Hirst, 1962) and Buzzards Bay, Massachusetts (Moore, 1963), chromium is
undoubtedly distributed with the clay fraction, but, in the present study area,
chromium is primarily related to accessory minerals.

Copper. The geochemical association of copper in Cardigan Bay deposits is a
complex one, and it appears that copper is related to phyllosilicates, accessory
minerals, even some sulphides in lithic fragments and in rare quartz grains. Weber
& Middleton (1961, p. 250) observed that copper varied in its geochemical associa-
tions, and they, too, found it difficult to assign a single host. Hirst (1962, p. 1167)
suggests illite (equivalent to our 10 A muscovite) as the probable mineral carrier.

120 SEDIMENTATION IN NORTHERN CARDIGAN BAY

Ivon. From a study of the data, it is concluded that iron in Cardigan Bay sedi-
ments is primarily related to the accessory mineral suite. Some iron is contained
in the mineral chlorite, but since this mineral is found in such small amounts, its
iron contribution is negligible.

Magnesium. Magnesium is concentrated in the phyllosilicate suite. Similar
findings by Weber & Middleton (1961, p. 250) and by Moore (1963, p. 548) confirm
the magnesium-clay relationship. It is suggested that magnesium is most abundant
in the 14 A chlorite fraction.

Manganese. It is evident from the reported data as well as Text-fig. 28 that
manganese is highly enriched in these shallow water marine sands. Such enrichment
is uncommon, for even where clays are abundant (Moore, 1963, p. 548 ; MacPherson,
1958, p. 78 ; Goldberg & Arrhenius, 1958, p. 170), the deposits are not as rich in
manganese (proportionate to the clay content) as are the Cardigan Bay sands. The
anomalously high values for this trace element appear to be related in some way
to the nearby rocks (Mohr, 1959), where rich manganese shales are exposed, particu-
larly in the Harlech dome area. Thus, manganese in these sands is related both to
lithic fragments and to accessory minerals, possibly including garnet. Weber &
Middleton (1961, p. 250) reported that, for rocks, some manganese is associated with
the carbonate fraction and some with heavy minerals. It is doubtful whether any
manganese is related to carbonate minerals in the Cardigan Bay deposits, for when
it is, the bioclastic debris is usually stained light brown ; a feature not observed
during sample study. Staining on shells from Tremadoc Bay has been observed,
and in this area a partial Mn/carbonate association may exist. Rao (1962) in a
study of manganese distribution in Recent sediments off the coast of India found
that manganese was concentrated in the shallow inshore sediments, and that its
highest concentrations were in silty clays. Furthermore, he found an antipathetic
relationship between manganese and calcium carbonate.

Nickel. Nickel is related to the accessories and the phyllosilicates. In the case
of certain samples, such as for St. 240 (145 ppm) and 67 (135 ppm), the high nickel
content is undoubtedly related to accessory minerals; however, for most other
enrichment zones, nickel is related to lithic fragments in the sediments. In either
case, it is a good indicator of sediment dispersal. The constancy of the Co/Ni ratio
for marine sediments, in general, has been previously discussed.

Lead. Lead in Cardigan Bay sediments is primarily related to the phyllosilicates
bound within lithic fragments, but for the area near Borth some lead may be related
to detrital fragments of vein quartz. The major dispersal of lead-bearing sediments
(Text-fig. 30) away from the Borth area was studied in detail, and a small amount
of lead is present which cannot be accounted for by the reported diffraction data,
or by the textural data (relative to the accessory/texture relationship). The only
conclusive evidence would be that obtained by electron probe analysis. Lead,
incidentally, is enriched in marsh sediments, but depleted in littoralsands. Turekian
& Wedepohl (1961) gave 7 ppm as the average lead value for sandstones. The
present data for Cardigan Bay, considering the relative mineralogy, suggest that the

SEDIMENTATION IN NORTHERN CARDIGAN BAY 121

lead content of these sands is slightly higher than their reported average. Some
clue to this enrichment may be found in the fact that the adjacent mid-Wales region
was once an active lead mining area. It must be emphasized, however, that “ con-
tamination lead ”’ is not being delivered to the bay. This may be verified by refer-
ence to the data for river samples. Hirst (1962, p. 1,168) reports lead values of
I3 ppm and 17 ppm as averages for the Gulf of Paria delta and platform sands.
His averages are in agreement with the present study for most of the fine sands in
Cardigan Bay.

Scandium. This element, like gallium, is related to the feldspars and phyllosilicates
where it replaces aluminium in the crystal lattice. It can be used, rewardingly, in
provenance studies.

Strontium. By far the greatest amount of strontium is concentrated with calcium
in calcite which forms the shell material of marine invertebrates. Some strontium
is present in river samples and outcrop samples which do not contain bioclastic
calcite. Nevertheless, strontium is still related to calcium, geochemically, even
though the host may be a phyllosilicate. Excepting where calcite values are parti-
cularly high, strontium values remain fairly constant for the inshore fine grained
sands with values between I00 and 150 ppm. Strontium is low in littoral and river
deposits.

Vanadium. Vanadium plotted with iron shows a positive correlation. Inasmuch
as most iron in Cardigan Bay sediments is contained in accessory minerals and
chlorite, we may assume that vanadium follows a similar geochemical pattern in its
distribution. Coarse lithic fragments cause anomalous variations at several stations
with poor sorting.

(7) Comments on economic applications

The correlation of known, or potential reservoir sands, is a formidable problem
in many oil producing provinces, particularly where faulted sections prohibit
“ straight line ” correlation, and where the use of electrical resistivity and radiometric
logging techniques do not differentiate to an adequate degree. It is possible that
local correlation, e.g., across faults, may be established by the use of the grain classi-
fication technique, as presented in this study. Conventional, indeed even sidewall
cores, would be of sufficient size to provide adequate sample material for such
thin-section analysis. The obvious problem, of course, in a new oil development
programme is that not enough bore holes may have been drilled to provide a coverage
equivalent to the present sample distribution plan. By plotting one quartz variety
against another, one could identify beds and, thus, establish local correlations on
the basis of scatter diagrams alone. Caution must be exercised in attempting more
than local correlation. On the other hand, the “ quartz approach ”’ applied to even
a large area is rewarding in provenance study and environmental reconstruction.

It must be recognized that textural and geochemical correlations (e.g., trace
element ratios) across wide areas may be more reliable than correlations based
solely on petrographic types, particularly where sample distribution is limited.

122 SEDIMENTATION IN NORTHERN CARDIGAN BAY

Nevertheless, in subsurface rocks which have been subjected to considerable faulting,
or which have limited sand facies development, the petrographic approach should
prove useful. Correlation by geochemical means, particularly the use of trace-
element ratios may prove worthwhile in certain circumstances where the mineralogi-
cal content is known. It is suggested that cobalt, gallium, nickel and copper would
prove suitable elements for undertaking geochemical correlation of arenaceous
strata. The data should be used to establish ratios for comparison, rather than to
attempt correlations based solely on element abundances. For those interested in
locating strand lines, or recognizing translittoral sands, such variables as calcium,
strontium and barium and certain of the quartz varieties, should prove particularly
rewarding in environmental-depositional reconstruction. Careful study of the
reported data will suggest several additional clues for stratigraphic application.
While correlation has been considered in terms of petroleum geology, it is not re-
stricted to such an economic stratigraphic application, for ‘‘ The basic principles
underlying the technique of differentiating or correlating strata by means of their
stable mineral components are essentially those fundamental to the science of
geology ”’ (Milner, 1962, vol. II, p. 372).

Another problem in petroleum exploration is the recognition of a potential reser-
voir sand. The two most important characteristics are that the sand should be
reasonably fine grained and wellsorted. Such aspects as trace chemistry, mineralogy
and other purely compositional properties are subordinate to the basic textural
requirements of a sand network and good sorting. The importance of sorting is
that the interstices should not be filled with finer grained detritus even if it occurs
in small volume, in order to develop the necessary porosity and permeability.
Cardigan Bay sands possess the textural attributes of a good reservoir sand. In
order to display the overall textural properties of size and sorting, frequency histo-
grams of the majority of the Cardigan Bay deposits have been plotted (Text-figs.
46, 47). In Text-fig. 46, the ¢ Mz values for all samples between 3:5 and —1°5
are plotted in the form of a simple histogram. Clearly, the vast majority of the
samples are fine sands. Inasmuch as the sample network is representative, this
figure shows at a glance, the sandy nature of the bay floor. The second and, perhaps,
more important quality, is that of good sorting or dispersion. Reference to Text-
fig. 47 shows that the Cardigan Bay sands are well sorted with the majority being
within I-10 Soand 1-30 So. Although the influence of secondary interstitial cementa-
tion and packing fabric cannot be predicted, the two prime requisites of uniform
fine sand and good sorting are recognized for Cardigan Bay. While it is not within
the scope of this study to discuss oil field reservoir lithologies, their recognition and
correlation, the results of this study do provide additional criteria for application
by exploration geologists.

Potter and Pettijohn (1963, pp. 14-15), in discussing the facies model state that

. a clastic dispersal system produces attribute, scalar and directional properties
forming an interrelated set that can be used to reconstruct the original conditions of
sedimentation. With adequate data and an understanding of the principles involved,
one should be able to make more successful predictions of sedimentary trends.”’
Certainly in the process of exploring for oil and gas, the competitive advantage is

ee

SEDIMENTATION IN NORTHERN CARDIGAN BAY 123
30

_

OF SAMPLES
nN
fe)

NUMBER
re)

=O) 1@) 1 2 3
> Mz

Fic. 46. Histogram of grain size values, Cardigan Bay sediments. The majority of the
samples are in the sand range, a desirable textural attribute of reservoir beds.

a

NUMBER OF SAMPLES

10 11 12 13 14
SORTING (DISPERSION) So

Fic. 47. Histogram of sorting values for Cardigan Bay sands. Note that low So values
predominate. A lithified sand with such sorting and with negligible cement would be a
desirable reservoir rock.

held by the operator who, often with very limited data, first predicts the trend
pattern of the reservoir. Indeed, the need to predict, accurately, the pattern of
clastic reservoirs, i.e., those controlled by sedimentation, is a prime requisite for
successful exploration. We have seen that Cardigan Bay dispersal salients are
chiefly related to the charted distribution of “reservoir type’”’ sands. Thus, in
attempting the prediction of sedimentary patterns in ancient sands, every effort
should be made to secure such comparable textural and geochemical data as will
provide a framework amenable to scalar and vector mapping.

124 SEDIMENTATION IN NORTHERN CARDIGAN BAY

While rigid mathematical treatment of dispersal data is a relatively easy task
with modern computers, the use of vector algebra, trend surface maps, and the like,
is not necessary if a simple graphic (vector) technique, such as used here, provides
an objective, representative presentation. The initial attempt in plotting dispersal
patterns should be the most straightforward one, often a simple graphic approach
will suffice, retaining the detailed computations for complex problems and obscure
relationships.

Moreover, the stratigrapher or economic geologist who wishes to apply the results
of Recent sediment investigations, such as the present study of Cardigan Bay sands,
should keep in mind that the studies of smaller areas are normally of much greater
resolution. Thus, they should be used as interpretive guides only in light of the
findings of other Recent sediment studies covering broad segments of depositional
basins or shelves in order to achieve the proper perspective. Such excellent studies
as those of van Andel (1960) and Koldewijn (1958) provide proportionately larger
coverage, particularly for broad facies relationships. Furthermore, modern analyti-
cal approaches developed during the past few years have provided the marine sedi-
mentologist with means which, when properly applied, allow him to differentiate
between relict and modern deposits.

In short, results of this work and other Recent sediment studies provide criteria
for interpreting ancient depositional environments; for predicting sedimentary
patterns where a minimum of stratigraphic control is available, and for establishing
local correlations.

IX. SUMMARY AND CONCLUSIONS

The results of this study suggest the following conclusions.

i. In the northern, sand-floored part of the area, bathymetry is controlled by
the local sedimentation of sands introduced from beyond Sarn Badrig and from
westwardly directed shore drift ; thus, net deposition over scour is responsible for
the bathymetric high or shoaling zone extending southwestwardly from the land-
ward end of Sarn Badrig. From this it may be concluded that, in shallow high
energy environments, such as the present study area, bathymetric highs are related
to sites of active transport and net deposition, and that some troughs, or bathy-
metric lows, are related to zones of scour or transport. In low energy environments,
these associations are frequently reversed.

2. In spite of mountainous topography and high annual rainfall for the adjacent
land, the majority of the detritus entering the bay is not derived from the nearby
drainage area, but rather it is eroded from the shore exposures by the surf during
winter gales, and to a lesser extent by normal beach waves during the remainder of
the year. There is no evidence in the accumulated data of the mountainous land-
mass or of the abundant rainfall.

3. From a combined study of bathymetric and compositional data, and limited
wave and current observations, it is concluded that tidal currents are the most
important agents of sediment dispersal in the bay.

4. Texturally, the greater part of the survey area is floored by well sorted fine
and medium grained sands ; there is no pronounced difference between a chart

SEDIMENTATION IN NORTHERN CARDIGAN BAY 125

based on contoured ¢ Mz measures and a chart based on simple facies grouping of
similar textural types using a ¢ Mz-based Wentworth nomenclature. In general,
the finer grained sands are located closer to shore while the coarser grained sands
are farther seaward. However, by charting the deposits according to the Niggli
scheme, bimodal sand and gravel deposits are recognized. Thus, certain deposits
classified as coarse sands, very coarse sands and granules by the single ¢ Mz measures
are, in fact, mixtures of fine sand and fine gravel.

5. Coarse grained, poorly sorted sediments distributed northward from the
seaward terminus of Sarn Wallog and mixed with fine sand suggest either a source
of sediment associated with the erosion of the sarn, or of sediment transported into
the area from the south of Sarn Wallog. Thus, the mixing of two sorted populations
results in pronounced textural dispersion, and such a finding may be used to define
the charted limits of polygenetic deposits, and aid in establishing their origin and
paths of transport.

6. Thin sections of Recent sediments permit an inclusive, empirical grain classi-
fication scheme to be employed, thus providing a firm basis for charting specific
grain type distributions and for classifying deposit types. This system is objective,
singularly descriptive, flexible, and the grain count data can be charted to determine
dispersal patterns.

7. The Pettijohn (1957, pp. 525-528) concept of dispersal shadows for clastic
sediments is applicable to the sediments in the northern portion of Cardigan Bay.
Thirteen grain types are good indicators of sediment origin and dispersal.

8. Photomicrographs of river samples for the Afon Dyfi show that the deposits
are composed of slate fragments, whereas samples from the Dyfi estuary/river
transition zone contain both fine grained quartz sand and fragments of both slates
and arenites. Such observation permits the conclusion that estuary sand is derived
from the bay—a conclusion which is also supported by geochemical evidence.

9g. Photomicrographs of thin sections of samples adjacent to the Borth cliffs and
samples from the bay floor show the presence of numerous grit fragments similar
to the Aberystwyth greywacke beds exposed nearby. It is concluded that most, if
not all, of the grit fragments in this bay area are derived from these coastal exposures.

ro. Increase in the phyllosilicate content of a high energy sediment is related to
the increase of lithic fragments alone.

Ir. Calcite in these high energy, nearshore sands is related to bioclastic debris
and is added to the sediment system only at the site of deposition.

rz. A composite dispersal plan for the northern portion of Cardigan Bay shows
that three important sources of sediment are within the survey area : the cliffs of
greywacke exposures near Borth, the low cliffs of glacial sediment exposed at Towyn,
and the coastal exposures of glacial sediments north of Sarn Bwch. Sediments
enter the survey area from without at the shoaling end of Sarn Badrig, near the
seaward terminus of Sarn Badrig, near the seaward terminus of Sarn Wallog, and
from beyond the western margin of the survey. The net transport of sand is north-
ward with a partially developed counter-clockwise system in both the north part
and south part of the survey. Moreover, the estuaries are being filled from the sea
and not from the rivers.

126 SEDIMENTATION IN NORTHERN CARDIGAN BAY

13. In general, the composite dispersal plan based on grain types is the better
indicator of sediment origin and distribution, while the element distribution plan is
the better indicator of tidal currents and most active detrital transport zones.

14. In the present study, the majority of the Cardigan Bay sands are classified
as sub-greywackes according to the van Andel scheme, but as quartzose sands
according to the Pettijohn scheme. The latter categorization is preferred for
Recent sediments.

15. The most mineralogically mature sediments are the fine sands with values
between 2-0 and 2-5 @ Mz. These deposits include the maximum amount of quartz
and the minimum amounts of both feldspars and phyllosilicates, and they are distri-
buted in the zones of most active tidal current transport. They are the best sorted
of the several textural groups, and while the total amount of quartz present in a
sample (X-ray determined) is related to texture, no evidence suggests textural con-
trol of quartz dispersal. This finding greatly enhances the value of quartz in
determining sedimentary dispersal.

16. Enrichment of zirconium in fine sands is related to zircon concentration (a
lag deposit) in areas of most active bottom currents. Moreover, zirconium values
exceeding about 400 or 500 ppm, in fine grained sands such as these, are indicative
of high energy sedimentation.

17. The relationship between titanium content and median diameter is also
related to the environment of sedimentation. Titanium is present in these deposits
largely in the accessory mineral suite, and, as such, it is controlled by both texture
and bottom currents. For most of these sediments with Md values below about
0-40 mm., and for all fine sands, titanium is present in the accessory suite ; however,
for the deposits with Md values exceeding about 0-70 mm., titanium is also present
in phyllosilicate minerals, chiefly within coarse slate fragments. Since coarse
lithic fragments normally reflect high energy just as accessory minerals do in areas
of fine sand alone, increased amounts of titanium in sediments such as these are
correlative with increased tidal current energy.

18. A characteristic relationship exists between the amount of gallium present
in a deposit and the sum total of phyllosilicates and feldspars. The amount of
gallium replacement as well as the slope of the gallium/phyllosilicate-plus-feldspar
regression does exhibit variations reflecting provenance control.

1g. The sharp increase of titanium with the relatively small increase in iron is
characteristic of high energy, well sorted sands, in which both iron and titanium
are related mainly to the accessory mineral suite. Increase in titanium content of
sand from approximately 1,000 ppm to 5,000 ppm with less than 3% increase
in total iron, for sediments such as these, is significant evidence of the high energy
nature of the depositional environment.

20. Increasing strontium with increasing calcium values is characteristic of
these marine deposits and others of a similar nature.

21. The slope of the Ti/Zr regression for sediments from any one environment of
deposition may be used to characterize the energy/sediment relation. Thus, sedi-
ments containing detrital clay and having an inverse Ti/Zr relationship characterize
the “ normal’ low energy site of deposition, whereas positive correlation is charac-

SEDIMENTATION IN NORTHERN CARDIGAN BAY 127

teristic of a high energy environment. There are, of course, intermediate types
between the two extremes.

22. The use of the scandium/gallium ratio aids in interpreting spectrochemical
data relative to the probable provenance. Scandium and gallium represent an
ideal example of the correlation of two trace elements which are genetically related
to each other and to the provenance area.

23. It has been observed in the present study that a more or less constant ratio
exists between cobalt and nickel. Thus, if the present study and others show
agreement, and if Turekian & Wedepohl (1961) are correct in saying that many
metamorphic rocks generally retain a trace element composition similar to their
unmetamorphosed equivalent, the Co/Ni ratio should be a useful interpretative
criterion for the study of metasediments.

24. Chemical elements present in Cardigan Bay sands are chiefly associated with
the following compositional groups :

(x) Alumino-silicates : Al, Ba, Ga, K, Mg, Na, Sc, and minor amounts of B, Fe,
Cr, Ti and V<

(2) Accessory minerals : B, Co, Cr, Cu, Fe, Mn, Ni, Pb, Ti, V and Zr, and minor
amounts of Pb and Sc ;

(3) Carbonate : Ca, Sr, and minor amounts of Mg, and possibly, Mn ; and

(4) Vein quartz : possibly, minor amounts of Cu and Pb.

25. Study of the petrographic data shows that associations of certain quartz
types are common over distances of several miles in the fine sand facies. Thus,
correlation of sub-surface sands should be possible over similar distances by analyzing
conventional bore hole samples using the same petrographic method. Such a
technique should prove useful in establishing local, but critical, correlations of sand
bodies associated with oil field development.

26. The sands in the northern portion of Cardigan Bay possess the desirable
sedimentary characteristics of a potential reservoir rock.

27. The results of this study suggest that the direction of dispersal and the
direction of depositional slope are not always the same. Such an observation could
be useful in interpreting the local paleogeography of ancient deposits.

28. The reported textural, compositional, and chemical data provide significant
clues to sediment origin and dispersal; thus, the data should prove useful in com-
parative studies of similar, but lithified, sediments in the geological column.

X. ACKNOWLEDGEMENTS

I am very grateful to Professor Alan Wood for his counsel during the course of
this research ; for providing facilities and equipment, and for making my academic
visit to Great Britain possible. To Dr. John D. H. Wiseman, I owe a debt of grati-
tude for his encouragement of this study ; for critically reviewing the manuscript,
and for his unfailing spirit of scientific co-operation. Thanks also go to Dr. John
Haynes and Dr. Hugh Jones for their help at sea, and to Dr. Harry Dahl, Dr. Wilbur
Hall, and the Management of Texaco Inc. for their assistance in the diffraction and
spectrochemical analyses. To my wife, Jerry, I am deeply appreciative for her

128 SEDIMENTATION IN NORTHERN CARDIGAN BAY

help and understanding during the course of this study. Without the generous
financial support of the Department of Scientific and Industrial Research, this
investigation could not have been undertaken. Lastly, I thank the Society of the
Sigma Xi for a grant-in-aid of this research.

XI. REFERENCES

Avams, T. D., Haynes, J. R. & WALKER, C.T. 1965. Boron in Holocene illites of the Dovey
Estuary, Wales, and its relationship to palaeosalinity in cyclothems. Sedimentology,
4: 189-195.

Bassett, D. A. 1961. Bibliography and index of geology and allied sciences for Wales and the
Welsh Bordevs, 1897-1958. National Mus. Wales, Cardiff.

Bratt, H. & Curistiz, J. M. 1963. Undulatory extinction in quartz of igneous and meta-
morphic rocks and its significance in provenance studies of sedimentary rocks. J. sedim.
Petrol, 33 : 559-579.

BLUNDELL, D. J., Kine, R. F. & Witson, C. D. V. 1964. Seismic investigations of the rocks
beneath the northern part of Cardigan Bay, Wales. Q. Jl geol. Soc. Lond. 120: 35-50.

BorisENoK, L. A. & Sauxov, A. A. 1960. Geochemical cycle of gallium. Int. geol. Congr.,
Norden, 21st Session, part 1 : 96-105.

Bromiey, A. V. 1964. Allanite in the Tan-y-grisiau microgranite, Merionethshire, North
Wales. Am. Miner. 49 : 1747-1752.

Carr, M. H. & Turextan, K. K. 1961. The geochemistry of cobalt. Geochim. cosmochim.
Acta, 23 : 9-60.

CHALLINOR, J. 1949. The origin of certain rock structures near Aberystwyth. Proc. Geol.
Ass. 60: 48-53. a

CorRENS, C. W. 1935. Die Sedimente des Aquatorialen Atlantischen Ozeans; Die verfahren
der gewinnung und unterssuchung der sedimente. Wiss. Ergebn. dt. atlant. Exped.
“ Meteor ’’, 3, pt. 3 : I-42.

Cox, A. H. 1925. The geology of the Cader Idris Range (Merioneth). Q. Jl geol. Soc. Lond.
81 : 539-594.

— & We ts, A. K. 1920. The lower Palaeozoic rocks of the Arthog-Dolgelly district,
Merionethshire. Q. Jl geol. Soc. Lond. 76 : 254-324.

1927. The geology of the Dolgelly district, Merionethshire. Proc. Geol. Ass. 38 : 265-318.

Fork, R. L. 1961. Petrology of sedimentary vocks. Hemphill’s, Austin, U.S.A.

GOLDBERG, E. D. & ARRHENIUS, G. O. S. 1958. Geochemistry of Pacific pelagic deposits.
Geochim. cosmochim. Acta, 13 : 153-212.

Grim, R. E. 1953. Clay mineralogy. McGraw-Hill, New York.

GRIPENBERG, S. 1934. A study of the sediments of the North Baltic and adjoining seas.
Fennia, 60: 3 : 1-231.

Haynes, J. R. 1964. Live and dead foraminifera between the sarns. Nature, Lond. 204: 774.

Hirst, D.M. 1962. The geochemistry of modern sediments from the Gulf of Paria, II. The
location and distribution of trace elements. Geochim. cosmochim. Acta, 26 : 1147-1187.

Hyjutstrom, F. 1939. Transportation of detritus by running water; im Trask, ed., Recent
marine sediments. Am. Ass. Petrol. Geol., Tulsa.

Hovueu, J.L. 1942. Sediments of Cape Cod Bay, Massachusetts. J. sedim. Petrol. 12 : 10-30.

HuckENHOLz, H. G. 1963. A contribution to the classification of sandstones. Geol. For.
Stockh. Forh. 85 : 156-172.

HypDROGRAPHIC DEPT. 10962. Ivish Sea pocket tidal stream atlas. The Lords Commissioners
of the Admiralty, London.

HYDROGRAPHIC OFFICE. 1934. Chart 4454. New Quay to Great Ormes Head. 6th edition,

from British Surveys to 1929. U.S. Navy, Washington.

1951. Publ. No. 145. Sailing directions for the west coasts of England and Wales. 4th

ed. U.S. Navy, Washington.

SEDIMENTATION IN NORTHERN CARDIGAN BAY 129

Inman, D. L. 1949. Sorting of sediments in the light of fluid mechanics. J. sedim. Petrol.

19 : 51-70.

1952. Measures for describing the size distribution of sediments. J. sedim. Petrol.

22 : 125-145.

JeuHu, R.M. 1926 The geology of the district around Towyn and Abergynolwyn (Merioneth).
Q. Jl geol. Soc. Lond. 82 : 465-489.

Jones, O. T. 1912. The geological structure of central Wales and adjoining regions. Q. JI.

geol. Soc. Lond. 68 : 328-344.

1956. The geological evolution of Wales and the adjacent regions. Q. Jl geol. Soc. Lond.

111 : 323-351.

— & Pucu, W. J. 1935. The geology of the districts around Machynlleth and Aberystwyth.
Proc. Geol. Ass, 46 : 247-300.

Keerine, W. 1881. The geology of central Wales. Q. Jl geol. Soc. Lond. 37: 141-177.

Kein, G. DE V. 1963. Analysis and review of sandstone classifications in North American
geological literature, 1940-1960. Bull. geol. Soc. Am. 74: 556-570.

KoLpEwIJN, B. W. 1958. Sediments of the Paria-Trinidad Shelf. Mouton & Co., Den Haag.

KrumBEIN, W. C. & ABERDEEN, E. 1937. The sediments of Barataria Bay. J. sedim.
Petrol. 7 : 3-17.

KrumMBEIN, W. C. & PetTIJOHN, F. J. 1938. Manual of sedimentary petrography. Appleton-
Century, New York.

KRYNINE, P. D. 1940. Petrology and genesis of the Third Bradford Sand. Bull. Miner.
Inds Exp. Sin Penn. St. Univ. 29.

Ler, A. 1960. Hydrographical observations in the Irish Sea, Jan.—March, 1953. Fishery
Invest., Lond., Ser. II, 23: 1-25.

MacPuerson, H. G. 1958. A chemical and petrographic study of Pre-Cambrian sediments.
Geochim. cosmochim. Acta. 14: 73-092.

Mat ey, C. A. & Witson, T. S. 1946. The Harlech Dome, north of the Barmouth estuary.
Q. Jl geol. Soc. Lond. 102 : 1-40.

Mitner, H. B. 1962. Sedimentary petrography, 4th ed., 2 vols. Geo. Allen, London.

Monr, P. A. 1959. A geochemical study of the shales of the Lower Cambrian Manganese
Shale Group of the Harlech Dome, North Wales. Geochim. cosmochim. Acta, 17 : 186-200.

Moore, J. R. 1963. Bottom sediment studies, Buzzards Bay, Massachusetts. J. sedim.
Petrol. 33 : 511-558.

— & Garraway, S.H. 1963. Use of polymethyl methacrylate in preparing thin sections
of Recent sediments. Nature, Lond. 200 : 62-63.

PETTIJOHN, F. J. 1949. Sedimentary rocks. ist ed. Harper & Bros., New York.

1957. Sedimentary vocks. 2nd ed. Harper & Bros., New York.

1963. Chemical composition of sandstones—excluding carbonate and volcanic sands.
Data of Geochemistry, 6th ed., Prof. Pap., U.S. geol. Surv. 440-S, pp. 1-19.

Porter, P. E., Suimp, N. F. & Wi1tTErRS, J. 1963. Trace elements in marine and fresh-water
argillaceous sediments. Geochim. cosmochim. Acta, 27 : 669-694.

Potter, P. E. & Pettijoun, F. J. 1963. Paleocurrents and basin analysis. Springer-Verlag,
Berlin.

Rankama, K. & Sauama, T. G. 1950. Geochemistvy. Univ. Press, Chicago.

Rao, M. S. 1962. Manganese in the shelf sediments off east coast of India. Pyoc. Indian
Acad. Sci. A, 56: 274-284.

RITTENHOUSE, G. 1943. The transportation and deposition of heavy minerals. Bull. geol.
Soc. Am. 54: 1725-1780.

SHAw, D. M. 1954. Trace elements in pelitic rocks. Bull. geol. Soc. Am. 65 : 1151-1182.

SHEPARD, F. P. 1954. Nomenclature based on sand-silt-clay ratios. J. sedim. Petrol. 24:
151-158.

——, PHLEGER, F. B. & vAN ANDEL, T. H., eds. 1960. Recent Sediments, northwest Gulf of
Mexico. Am. Ass. Petrol. Geol., Tulsa.

“wy pod 3 uh 9!

‘Seuchtcai ha 16, © ihe dilery eal
os fa alin. Paden si #0 ee » 95
bran, PO Chi a 8 «the ell ies of th: ie eabtaaidel a v
ii atted Steen: -.. o>:-flaye Oy omar M4 = on” 4? i - =

‘Steen, A iis a “weeds enbcthiovs cmos cht vilgdtvews abel? of Coreat 8
4 ; Late, ory hse : Aes * “y 6 i >
Tat “Tt 2 ' p Fide, te evs, @tr.. jplating Sa 1 ae logy, @iteralowy ws) t
m av of Welew: 4 Te b, ley FP. Coneewe, ot Lael i od eek A Tanah, Ree

4 ’ + - - * Zz ‘
. Terk é KA ewes 8 inp TOG ee ate

toe ee os
doite of ia Zari. Afvssal

a Lan yy try wink’ yA an ral a erg _ Gevukiaa:
wal eed) ho smi rebe ve ahead ikt? on ot haoeher zi bite eerovlint ot avvly af
Tor =. 4) ' x n : 2-3 kaes ai —“v be sakpur aya. plu
3 idly bit wll Pi Mripeiant » rng i oo boead in
VAR, 4 “ , sy is ay... FF mi ur gee

Sse Late ot ee ace cama

-Aoatherit ovale b pres i. hatin wadiour ol aeg IAIN ide te berks ;
«alt “y i vis siiatip ' } 0) . holo . .B OL ae a > ew te * é ca
ad 15) , peel OMe fier * A cur 4, 7 it "hh vy ‘Pi tealdals batty @ PS =
Lei thhel 4 {4 9] -

Lh ae he UR 2 iar a oolaayt sth 28 Joyig iy TOMAR in,

4 of posmvane, bo tixak peed syad bas euley a stingray STs ewlev opal

i pi a ee | } wage ol pi te To}

ie eyourasnag frstifacty vor ALL fereedhipes a & fixh : <9? sn my vitae ig fe
e

ii Batowagig be TRO Nsipse b wis ‘SFR thi of ST oe DEH d

ghufin ignasal too sulle see Lin paHta SOS. TU (EO

1Ag isl Ry art te | : » aleve bit Sets FOUGINS amy? tec ons

Meswency ty cP ects Hie aewe teary Read aah: ye dovetabicteit boesit ps
‘ erevoraert we i btw herela irk elt) teal 30 |

rt inimota Nea foe) Gore rey aq mt | NI TALST x1 “Pauley bs a
ab ve iF ong a) (Ou9 vi es) crt fit ;

TABLE III
(St. 1-57)

Spectrographic

Chemical Elements,

Optical Potrography, Grain Types, Percentages

ancum 2982 QEESY S2EH9 BRSSR SHSSH BSSSE SESS] SShoy LESS os
ORE) Choc a ii a i =
Vendun 9885 SS9h@ Sala TESTS SSSTS SSRal FSSSS SSSR SSRI Rs
esese e22ss 28882 $8288 S288 SESS SESk Sess SSSss ss
Titanium 22888 Seh22 SESS BSES2S oN TRa SSES2 S5222 SESE Gases SF
acim BESS S883 SESEL SESE ESEES ESSER MEh=S Sesh SESeS oo
comes, «SC UBESE SEELEY SRSS2 Soeee) Ress Sees Shes Shoes Sees co
Lead Bess Sar
Mogimre Sas52 S2058 Swans GOene Beas2 22225 Baess S525 SSess a6
Gallium Si
Iron Ss
Copper es
Chromium os
Cobalt 25
Calcium ah
Borium Se
Beron a
Aluminium ans
Foldspara aa
Cher ae
Crystallines S 3
Shale-slate aS
Arenites Ba
Carbonatoz as
Quartz 16 ae
Quartz 15 ce
Quarts 14 de
Quartz 13 ae
Quartz 12 Sey
Quartz 1 za
fe
Quortz 10 en
Quartz 9 aie
Quartz 8 Ss
Quanz 7 on
Quartz 6 20
; od
Quanz 5 ne
Quartz 4 w=
Quartz 3 on
Quanz 2 ae
Quartz 1 =o
Peldspars BIND D=oar BAVSS Broan FeOeD BwBoKT BRHOS ScoTTo BOEn”D wa

Phyllosilicates means nrnoo Roane oate. @rene Sanam mueoe onan

2) pelomite PMSS4 SNPS BSotm essse ceges poEes GoSoa oadem Samco os
2 Calcite DOGS ARORA WSMUS SVBHo SHEhO SHGan EHMSS GERoD BORG a
SALI ocMEm GaAMS momMM HoTSa BamMOS momo
> Ottheclase nen ingame oman omawn ametne oe==n ao
*% Quartz RESSS slVee 2Seece sesee venn= weenn sone. zanao e=g 39
Chlorlte ATAPI Nine T Aen aan Baden TeRNN Nene von
Mica SisiniS ANNI Nin) SoNimio Saw wan Sei one ==
Background «ieee mxren wesre zeeron zones one
= Dolomite “head mets «anes ams
2 > Tt ATO Ned are wana om
5 Calclta SE2e@e noes mwanmn wena= =sa-0 a
5 Seo =a0
U Plagicclase SN OnG nanan antom aonman nooan
G CUES QOOGH Q2vEH OHOMS VEGA, ONSoOo Hea GoOooR HaMAE Gaon :
© Quartz REk2 geane aa
: RESER SSSRE SRRES SESE BEES =-oOS Seqao Reema Roem oo
& RS SSRRL TkGes SLR RARSS Sears ss
= Chiorite SR OMe wronwe Rraem oO
Mica SOM ane

Mz SQS52 2280s nee ante seeen senna seman ange nean= an
3 p84 SQGSs Saees meen nreams gaaea mpmees @nenn Mune nneoe ow
= OMs0 SSS SNe ace Sole meciein nudia= Mac<n| same ning) «.
a
aaa GUISES Geage S2ahim GEES Gamam GENRE Gamam aaaee maces =m
2 sk SSens SS"S5 Seas sees sSene seage BSeee sakze 29se5 24
bj G23g4 2ZZRR BRTRS SEISBE GGG Soe goes FAyeh Geges es
8 S= deadd tage se-== Stet eaas= =e

So 3 S5225 S888 Sznen a=8g4 Beene euwes Sagn Araee an

Ma Sates SEghR SSS55 SESS8 GReSo Seeae GSSeq) HESRS Soaaq se

Sccec Sdoocc e009 Scena Sscc6cG cocoa Scoeca ccoce sacos oa

03 ates Geass Soke GROG SORE SSRs RoSss Gee Soke oe

©2 8090900 oo6cG scocca Seec0 SccocG GoceG ScoocG Sacce co

9 S8dco Gddce sddca dédad sédce duces S388 Qa89e IIsss gs
Depth eo
Station we

ete ce siesta

e- pri ake ates

35° ac 28%

at

eT 5

‘ ine ar)

é
| oe

: - ow
“aee-
7a > .

«28 ¢
eqeeoee
= = 6s
SsEcS
ee =

= ed
2 RO Ss,
a2 -
Z2eslea

a2 @
oe © ee
az -S6-
eat 6 «
ae ar
soe et
wera @ 6S
> Schell
gases
ones
a =
~stae

CPT a
a, =) we
‘2 rs
a ee
” ——
nl
‘
«
> «

to——a.9

OOF..t Oe oe

£
A
it

|
a
e

a
a
a

TABLE IV
(St. 58-117)

Chemical Eloments, Spectrographic

Percontages

Groin Types,

Optical Petrography,

Mineral Percentages

X-ray

Diffractogram Values

Basic X-ray

Textural Statistics

geese 2e2se SE5Sh $2282 ZS2es SESE2 sgess LSxSS S=ENas

Zeon BRGSS HSERR RSE 25 Be
oe peo Gonce syene ghee eanns
SSESS SS5S8 TSSS5 255 S2S8S SS8n8 SSasG SeeosS saans

Vensdun 5855 $5588 72555 22 S252

sesss e282 22388 SSSEE SEES SSESS FSSSE ZSZSE 223
S8sse Sasss seses S3ess SHss= SEERE SSHRE S=ESss SESSS
Heme RZanS SSS52 FR2S2 Fae== F=R2= S=RSR FSNNK KH eoe Now
Gapb Sess aroun cones ssns4 SESS S22S8 SESES
Stoning S222 asa Nec H Haake RBase Secoe Sa=e= —-=== ===s5

Sandia Sb Ges SfGad wkoes Reewo ~~od~ Wann aesan were = AREAe
Leod =e8) S22R2
Nickel SLES SSSS=
S58ca s2es8

Sodium Scocc do-ca
vege SHSS5 GSES Geces SSsRS SESS RESRS SRseS

Mangonese 23325 Ss7ce SSoen SSsss Nacne asthe sSms Some
Seec= $2335 39983 $8858 S4S85 55888 Sasa5 Saens

Magnesium = 3SSS6 Scces Gesec Saeco scccos sccoce soccs so-so
22025 SESS58 $528 S84Sa SSSFF Gahas angus sesss

Poassium = SS ac6 Soces Socce Ssoccs Sédco Sédcc Sédcc so-cs
Gallium $a5a5 Geen] Sdddel Guddal edden adcud Nades meteor

Tron cia

Copper Seles SRESS SSG5n Gotta seses Sasha eosos Sess Se eae
Chenun «35225 q2Rh2 BSTSS SFRS8 FSSRR FSLH5 SSRSA GRSAL SRSS=

Reobal Gesea Gets Tbr se Serre HOUSE LEUNG HeNee Fess aerSH

Calelum: Sdididd cindnn icinndd dnnan mata aad anna ann Sn

eznaae onmoy owo0e cecuy seone seees susseo wosue sesso
eesti Seon" Snead AnoNm Fosry enews enser Seoer enka atks=
BEG) SSnGS BSASGR S48a5 SAIAS Seaza Sonom Anaan AMT mow

bromn o=tam S@uen OnmT@ wrota Raat YNeNo Bonar arena
Aluminum ee ee edd Gdddd dddnin onddd Gaadd avdus adnan ansaid
Feldspara SaGs2 Gddan NGded Read= denae Nowa weoed aetans Sooon

Chert peeee Gatla dédea 246-6 =ccc= c--o4 =coc- ccceaa sca-—

Crystalline Sece= Gecad getaa 266-6 acon sand tana Soces otcan

Shals-slete EGG Gunga toted dua Saas wndne MNewn Noam soda
(ESISQA GIBBS SVQSS QDYSSSE BSwQo EQS) SOs woe iS
Aronltes GGnna Gidge weede eevee nédee qgond aang neta jason
Corbonetes MG gan Ekuda aennw tHtad anes aones Gana nae
SQQ2Q QIIAE QHVsO BESQVQ BYGSS Os999 GHSQs SBBeQ Qyeee
QuitzI6 dads Gecce ddc=c sedda dteca S666 Seccdc Scccc Séccc
GIIRS QIVay GQsep SI99G BEBeEsS Sagwey Gyuosa Ganeg geocs
Quitels G22SS Séddco acec= dadda Gdece Sodes S-dcc Secds Saccc
ABQIAY QBPVOQ QIQeS Qoggq SSang Qagqse wayNe Gaga agaes
Qunz So6Gc Gccce Geode daoc- sconce Séccs dcdcdc Séecc s4dcc
Quartz 13
Quarts 12
Quartz I
Quartz 10
Quartz 9
Quartz 8
Quartz 7
Quart: 6
Quartz S
NeTyR SReRe NoEAD KoAB ATeAD VanAy BREWIN BAQeN Enns
Qube Sd40 dde46 46nd tect oaddd v+c66 dacs +cceq scdcc
NAMEN =QeET Neene mens raw TenEN CaNN= RESON AMINe
Quatz3 Gann geodon “d=d6 22=6- dadao onan daton wien coded
griee Gat SES) GReY Gove cee Uecg Gude Gaeoa Gags
quasi 28252 SE255 SElSE Ge2SE FGEL2 Goede L2a02 FRREZ 2teE3
Carbonstes on nn nh Runnn weno Rreny neThy ennes Besar worge sa0re
Foldspars BeweEr Rvene wreDs BEN BRASS SSBES ArOBe HERaN Brine
Dolomite ee ee eee ee
Galette TOG SyoES RHOMn Sane SHnee UeTHH SYnRe nHoee RRO
Plagioclase mama ramet, corr nwo NTT tare SAToT TTR RII
Orthoclaze FMAM ANNA OFAN TAOMOM TOTNO NOOTT OTeNN ANTM =MI=H
ued BE8so SSSSs SLLSS SIsSS SSsSS FSS5S KSRLS KSSKR seege
Chlorite AINGIN MANO NNN Wann Wao WoNO Re wees NAOT ena
Mica NINN OD HNN ON NOK NON NNN NNT ONIN Oe oN
hegeant SQS3h SasQ0 Saeny Bawsy Seaee Beams Seen gogea BSgRqe
Dolomite ino maar TNO 9eT ON NONGN annme Swen tron swoow
Calcite Rane Shes nSnee anes g0e05 eraan Sean wean= Youre
Plagioclaze Gan yy Bmoey weroM Tyrer TeONH BHeE= NeeQe mrwen neARN
Ontheclass ao enm weHen wnaTs neoee YoETy 922Er wRaee ORR] —Hone
Quartz RSERG REEES RSSTR RSKRS RSRAS RSkks SARSS KELLL Heese
Chioite nto Memes TARA wHTON TMM TeOHT ewHSe Hrne® gangs
Mt GIBRE BEVIWS SswSS Sasdg Gageaq aaoye gages
dad Godda “Gd—=0 cdacdnd ddan adda Sogged
genes neo" veess wnyen evens ueugy uueee
dog guys SSSos SSSES SSeS HReas Sesee Saous
Sé-6e Sades Sodoc Gocde sddnc ceces Amine
Mz YSSAB BQPQD QISSSE BVMWD BaSSS SEAGy BVISPS sQsaea gqage
084 RROD GSGSe SIV SIRES VIII VOSIS Baa Sarge VE
M50 BAUER BEDS BOOAY Caso Gann qOBOA S =
Pls FEPDE PEWSY GRAS Gasag Ssqes EEGs SSUES
CSkRs S24RR FESS REESE EBRSTR BBRET S255
Sk ZPS3S SESS PEEES GUESS FSeye Hees Becee
T>+=8= Sseecs SoHo S==== =-=-6 saccHs eonoco
So S2SSs YEUVT GaSe SSeS Sees seus ea
Md Saas Guia GeWOS Seees NWek Waeee Bae
S6ced Sceds Sdoas dadec sdada oddaa BAESS
93 FSSS2 SURRY SSNS GOSS SEeS Hees Bang
83566 dcade ddcac dedda ddeda ddcad sta
AMUN QSNSR aS_ch Seeeg abeze Beene 5
cy) SSS HORI BGacy gcc Based Gohic Peeen
Séécé Sdddd dctcc dddde dcdcdc doded sages
Depth SIPQeE WSIS BIGGS SIGS BVSQS WEIeg ages Geog Sagas
Station SSScCS BESS SSnNR XRERR LSaSS SESSA FARR RSsSS sssee

560
580

0.59 0.53

0.33 0.38

13
3B

54
44

122) 10

4.56 0.45 -2.8

1.17

eet se SET Oey Mutat Fe hete

ae

yf

~
-

3

eee

m 7,
*

?-—
Pe eo
r ora

od
-_

<
“
~:

- *
e *

@aeca.s e096

22°28 gB4es
sees ee

*

-_sa @« &

on 2 Bed
_-~._sem
Fen «wi
»

21908 >

*

+ Sate T2555 TSP ve

dated
S355" $4345 3°4835

i ...

~
a

ma

“+

aA eS
- ~ .
oo o- i we
—— a
” “Te ee
t-
_
—

wr

|= wr ~
oa at

ay

445,

eer Ce
a5 +3535

“_« »
or Fh ht

arr Se

> hi A

-_
-

ene =
=

.
St et
“ene
=~" °

——

~—

’

a Ft a

=<
a

sehe

.
sae
iA

“a

x

‘ad

Wo

roe
ARES

a”
i")
—a ¢

<<

’
ie

ea ie

eo
”
*

=

=*

26

pay
s =ees3)

22553

&
*

ae
6 ‘i
‘ol “wy

va} tee k *éf,

OEE hs re 46;
Se * any
@ ..f°" wi
w tet * i.
mle "a, 4 a.
eiieehay.@* 0,
Ph ee i
a wre.
ce op, Pay
pales ® ra pts
le a

TABLE V
(St. 118-163)

Zirconium

Vanedium

Titanium

Strontium

Scandium

Lesd

Nickel

Sodium

Mangonose

Magnesium

Potassium

Gallium

Tron

Copper

Chomical Elements, Spoctrographic

Chromium

Cobalt

Calcium

Barium

Boron,

Aluminium

Foldspors

Chert

Crystallines

Shalo-slate

Arenitos:

Carbonates

Quartz 16

Percentages

Quartz 15

Quartz 14

Quartz 13

Groin Types,

Quartz 12

Quartz 11

Quartz 10

Quartz 9

Quartz 8

Optical Petrography.

Quartz 7

Quartz 6

Quartz 5

Quartz 4

Quartz 3

Quartz 2

Quartz |

Carbonates

Feldsparz

Dolomite

Calcite

Plagloclaze

Ortheclane

X-ray Mineral Percentages

Quartz

Chlerite

Mica

Background

Dolomite

Plogioclase

Orthoclase

Quartz

Chlorite

Basic X-ray Diffractogram Values

Mica

OMI

Gag

OMz

084

1OM50,

pe

Textural Statistics

So

Md

Q3

Qt

Dopth

Station

275

0.27 300
365

0.42 440
0,27
39 0.31
0.33 0,26 295

a

32
0.41
0.

0,26 270
1.06 430
0.32 286
1.1B 575
0,36 265

Gece
S2s23
ecoco

290
820

0.34 0,28

0,33 0,32 330
0.34 310
0.64 305
1.45

0.49
0.79
2.26

Zécco
ssses
=ccco
scocs

320

0,80 0.46 425
220

0,45 335
0.46 0.45 450
0,29 0.26

0.24 0.27
0.90

1,06 480

1,20 0,93 495

0.96

0,44 0,33 405
0.392 310

0.35
0,28

340

0.34

Sotee
Sec
Sots6

scc--
Scoc-

1,17 800
0.44 360

Phyllosilicates 4 on won

al
a5
80
60
86

3
rf
ri
4
rr)

73
73

77
71
78

nid.

n.d.

+ouan

74
78
60
65

rr)
18

wecon

0.13
0.24
0.21

0.26 0.20 0.23
0.22 0.18

0.30 0.13
0.25 0.18

Soca
Sszge
Seccca
=s254
ééccaca

0.17 1.06 0.99 2.3

0.16

0.18

132

0-10 0.12

0.14

eccco

2000e

147

148

las
150
151

182

57,
14

= 3y2es
era tc
Fetien 2&9

by ny we ge °

= . ie “ae ©

ae = eaonseq ceooe > had

- , =? +s w « - £
ow * cn z
Pe ee SSeS ww :
7 “ase pe ~~ eho > f
2ee8ts® S644. 808 SGeEeee #2! — f
— : ary Pe ee fn 8 ae
~ ard ed: “bao
gwies: 2s ete

——t —< r -
~~ wi eS 3 oom & ckbct a “ae 7
“ar, Ore = 4

$t-
* t-
J
i
'

t cela au -e
ad = / i ‘ ei
ae me 64 ot ee ee “me aS mw

s oie 2 bee 8 Se a i Se eS fore . m ~~
SOG5s 2207" 2255s +4.
- :
oe Ssae@, “ & - eww c=
>. he ee ee Se eae
‘ * 2
o — a2 ae, i OS OP ~ as +4 GS ao
»,9 + = * 4 * - *

-eeac™= cn & «& “~~ > ao ~
ip@aecme ecee<s aes S£88— ~ooS2 7-8~% «Eee €oe.v¥2e teeeu «se
“=! f vs 4% © ‘o ae a. . *- s ~ 2s " > * -—- ae 5» @ + * 6.2 - -

aee »- — a ; ond = a 6 - , > “use mw o os
38 2s SStss SSESes 2Se S2t¥E Per er! 2Saea 3 eee 258% ce ee
: ; : 7% 4
as -@ P x iy ¢
Lima EP A et Or 5 a er teter ee ee 2 = fae eo ao aad a Ee OR > As a wee
a + * ».93 * * aa * 8 * ° ¥ ° » * oo. #2 # . . - . > * - . . * - «
A kh eet ée2adea Tae @ « au 7 * 2«aee €£2252%90 *4£ &2= - 2. e s+ anv: 2

ks
+

ri
ae
pt
ik
4t
ot)
wt
b
i
.
¢
zs
|
e
-
r
#
g
i
§
1 a
b
b
'
"
4 a
é
x
f
. >
‘
:§
ti
y
at

bd : 4 « =
J ‘ 4
= ad Pi =~ oe = Wag -beked -- =s 7 2 s
ad a acccue a » s a = ol # & ’ ~~ -—— Ow oe a =“ © ~ ea © > * 2 >
2 See = “v7 e= ¥ Oo 4 2 7 “ ¢@
eS a = = seo j = 7
a ¢ ye ‘
& err ry & ee ee ee al cena at ~~ = wa a ie YS @ o- “<a a 4 @ Ts.) " ou w +4 <a
a = ‘er-=8 ©) o~ 4 ed ed ee 2+ we “wa rie “erw 4 -
_ er - - a | ie) > &- =~ ~ _ ci
>. : . —% = ~~ ve z
e ae
* eucens; te Soo ao a2 = =o + aawae on 2 es ifgwt) = ~ aS a 2
> ~ & %. ¥ = 28 =
~
. - “
al ’ — — —< — — =e
NT ret ~-Ss2e & os o> he et Fa S.-Sa “Ti«e= =
ga in ae gprs S
> Eier® Aas es '

TABLE VI
(St. 164-226)

Zircontum

Vanedium

Titanium

Strontium

Scandium

Lead

Nickel

Sodium

Spectrographic

Potassium

Gallium

Iron

Copper

Chemical Elamont

Chromium

Cobalt

Calcium

Barium

Boron

Aluminium

Foldspars

Chert

Crystalline

Shales-slate

Aronites

Quartz 16

Quartz 15

Percentages

Quartz 14

Quartz 13

Quartz 12

Grain Types,

Quartz 1

Quartz 10

Quartz 9

Quartz 8

Quartz 7

Optical Petrography,

Quartz 6

Quartz 5

Quarts 4

Quartz 3

Quortz 2

Quartz 1

Carbonates

Foldspars

Dolomite

Calcite

Mineral Percentages

Orthoclaso

X-ray

Quortz

Chiorite

Mica

Background

Dolomite

Calcite

Quartz

Chlorite

Basic X-ray Diffractogram Values

Mica

gui

dog

OMz

Cre

@Ms0

oi6

Sk

Textural Statistics

Md

93

Q

Depth

Station

106
120
266
86
120
140
139
179
620
275

109
We
55
37
85
98
113
78
a7
ch

5150
5250
2800
890
3650
4600
5000
3650
1850
2760

i
109
15
110
162
a2
100
127
122
450

0.6
31.0
8.0
3.8
22.0
20.7
44
13.0

4
27

196 24,2
W

167 34.0

680 3
63
7
Ww
48

56.0
56.0
16.0
8
33.0
46.0
49,0
40.0
1.6
27.0

1.17
1.07
0.15
0.17
0.76
1.10
tod
1,08
0.15
2.00

330
380
0,33 305
670
1.10 500
310
330

0.93 490
0.32

268

0.25 0.35 305
0.65 0.46

520

Manganese

1.30 1030
164

0.45 450
0,28 330

0.25 260

0,97

Magnesium

1,96 0.72 820

1.67
0.23

0.21
0.80 0.70 600

0.40 0.28 450
0.34 0,29 410
0,39 0,30 320

0.21
0.40, 0.31
0.28 0.31
0.34 0.34
0.21

Carbonates

Phyllosilicates

19

Plogloclase

80
76
80
82
83
42
34
50
86
46

27
32
24

1
18

21
24
13

n
W

15
V7
16
18
7
7
v7
16
7
16

26

26

24
15
19

Plagioclaze

Orthoclase

76
80
79
76
a0

10
12
7
9
7

2.4

0.20

1.98

1.09 0.99
1.26 1,06
5.59 0.15 -:
1.54 0.92 -3.3
2.34 1.01

370
470

0.28 0,36 480
370

0.72 460

0.46 0.50

0,26 0,42
0.37

0.56
0.21

4,55 B.A

9.2 40 15.4

3.38

290

14.

0.9

0.3

5

1

22.4

rT

82
78
61

4

W

18
20

10
1s

16

44
76
7
75
68
a
55

18
6
10
8
2
4

+.
i 7
"
¥ om 6966 1%
oF a
#% :
af. 3
*7 ip
4 *
2 8
sm «(tei
oi ' 8a
gv '?
su 48
aly Lh)
f ud
§ ii
e ¢.}
eg e
t é
p , ?
a 4
’ ’
an -
g wu
a Wa
% ¢
« 2"
‘ 2t
%
2’ ’
1
ti er
‘ hs
é
aj
of
i :
3 6
e
©
4 ¥
z
¥
’
‘ ‘
”

"=e = &

she!
o@

i”

a |

—p 9
geegm

¥
~~

,
- : 4 ;

- 28
Ss

e
=

vit Ay» hho

- aw > A nh a ae ide ns a
i se 6OF 2.9 Oe i Wola Sely Yon Made
Pe ee Oe ee ee eo re be
09 Oe GB » de ig ay 4
Ee OS ee 8b Fee nee tigd, 6244, 70,8, rH = Sy 4
cd 0.8 16 14 ES Gly et sin Way ears na
ht tp 2% 2h 8b OR bide af, ni ,
13 the BG 4a EG fete TO oe st
t@ p38. ¢€¢ 49 wt, echt
44 #4 \ee 3 is tes
je ws Vie 5 8h ee HA ied ts § 4
is wnt Ce *.5 64 ty td, th, "

‘4 640 t¢2 + i. Sly % 1 Ms t wce : ch b
| sa — , + Fe 4,! ~ ie oh A, ? g

; f i " Pp oe oP yal, oe 2 Bie os Uy

Va 7 am iad
ts a be Hi-~ 1H’ OO BS Pht <, &s
tid O@- tf t- hy OOH 8 \, BY
‘a--8¢ te le «ddr bl ee S bly.
bt Oe be 82 22. 26 Oe Be Gh 2h
ise 5 Ce - tc- ade &G 7¢ wl re
~ ee ie th) 1 Ok Oe ee ES

te ta e SF T6 He HS 8S

5 thm £3 c Ft wy te) = ie 5
ém t ,4. !# Lad io Site Sh.) iy i, +, 4,

; we 6 BA ts ki ad <5 th 4

(as

tii~ 2070 Bob Gee Gel- Geir ly Wd ee tel
- Oo, bt Ca Gf- Ad oO FES * ye
anb 6 1.2 ean at Ab Ah Gl 64;
43 d ».f te ot ey & .Y a! #4 “Why! :
mi tt ta he te OME UE GS
tots eEy4 ive 7c af) 6k AS OS OR bb o>
ia PAyt t.4 oe ho oe. 4 §%9 che S
ei! oe eee © ee
ys ate ie th tt BS Che BS EG ‘4
mi cho oR ha OS A OR A OG B®
ae tet Oe ee OG ee Oe
Mit 4Ols Mme Oe Sole ee Ti) as forte”

¢ £60 C4 teh Gh Bb ME ol TR HLS

f oe €2 ict &! a %.9 84. © 2 | =
ee Le ee ee. ee 6.9 4 . 4.5
wt Othe Be Vek Oe ohD By “) “Ee.
Ho 2 £8 yk -Set 4 44 ae +} j ¢ 3 4 .
%o Ol Fe i ti he oe ae oa
o> 0 fA doh: PAA ‘ 6 .f ae Hd ie
an ont ie Ce ke iS ee + wh. fe
.5- Byte tek Bite etl Oe aR AOE”

if
~ ie ; Ba. H rt
Vs : ish Ay . oy vA

TABLE VII
(St. 227-277)

2irconkum

Vanadium

Titanium

Strontium

Scandium

Lead

Nickel

Sodium

Manganese

Magnesium

Gallium

Iron

Copper

Chemical Elements, Spectrographic

Chromium

Cobalt

Calcium

Borlum

Boron

Aluminium.

Faldspars

Chert

Crystallines

400
440
450
620

0,19 0.37 445

0.23 0.28
0.52 0.43

0,22 0.34

Quartz 16

Quartz 15

Quartz 14

Quartz 13

Quortz 12

Quartz I

Quartz 10

Quartz 9

Quartz 8

Quartz 7

Optical Petrography, Grain Types, Percentages

Quartz 6

Quartz 5

Quartz 4

Quartz 3

Quartz 2

Quartz 1

Carbonates

Feldspara

Dolomite

Calcite

Orthoclase

X-ray Mineral Percentages

Quartz,

Chlorite

Mica

Background

Dolomite

Calcite

Plagioclaze

Orthoclase

Quartz

Chlorite

Basic X-ray Diffractogram Values

Mica

OMI

OM=z

aa

M50

Bis

Sk

Textural Statistics

So

Md

93

or

Depth

Station

Plagioclase

Phyllosilicates © 7 ~ or =

81
84
84
82
75

7
16
7
16
1g

anna

Ronne

Scocco
Seeae
Sodcaa
aence
Sees al
Sdddd

s
)
)
0
oO

8.
CH
10.1
8.
3.

227
229
230
231
232

0.26 0.17

0
0.22 0,36
0.35 0,31
0.31 0.43

daad

87
aa
as
a2
a2

i
7
v7
7
18

0.26 0.15
0.27 0.17
0.27 0.16
0.20 0.13

56
67
51
74

0.66 0.15

S253
S258
eee
eeees
Sgese

410
560

1560

0.29 370
0.27 0.33 515

0.45 0.45
0.31

5.4

3.65
2,50 5.4

79
81
7
a0
84

78

0.22 0.15
0.18 0.14
0.26 0.18
330

SESSe

eocoo

83
cht
a2
83
a

8)
17
72
78
61

0.17

0.21
21

1.07

1.69

6690095 ScoccG

08
1,08

1,26 0.94
1.35 0.98
0 i
19 3,22 ~
1
1.29 0.95
1

1.34

tiie?

eer & &
2 4—- @ eas ® -i

t

a4 =

ww

~

wre

i»

ne oe

ee
Bo)
jb.
5, Hi
Lb.aa,

ell atte
hyvtte 40,6
b.a,.@2) "> 46.4

Ls

*
c

ty
#

«
we
area yr &

fe wey eo aoe Ni on em

bit

-

aT
= i
. o-

pear =e)

|
wr

-
™
-
*

*

—-

ve i hehe os
oe Pa >;
Ss ang
oe.

s3

o ues oe
Seeke

a

asveéee =

ghee =
ginee-

i=]
+

une
‘= «

7m oe “seeo= rr = ites has
Pai) /
neg ©

&S ost

= eS

Sh BhOLE
Re a

» ax

7 7
- a oe int

cADRS

aKePE

4 hs 5a chat,
ESSN ott at, OOM! &

ah. Spm eS.

re

mn
ctnone

«

‘ em i wtb asd
a ee

a) i) =! at

on p me
ew «
en

‘iinetrad ]
v4 top 2.00 NS

et.ie ne 7, oe

“ *

; - wee

3-~ An oat gt

»

Qds Seles i Why
iam oes
ote Le a a

cm 7. Wt ie ar

anevwrr

es a
-—
=?

apkee 2nSE#

ake

iss
Sores

ha OF
crnn
Se)

. s

-

6

———

7+

“
-_

mn

—)
my

rd
Fe

-
a)
me ee Om,
Ca &

ek ee

w/e ae =
-oo
4

be os Ge
+2 -o ft
cess Keret

» <
PPS acter
s

;

~
a |
>

<a8ce C6

~

‘> weerd'

ro
ses
he

= ’ . .
rem =
~~
“==
~ =
=

man a

9 @
=
-
~
2
2D «-

oy

Pay Ais 2

Fic. 1. View of Aberystwyth greywackes exposed on the shore of Cardigan Bay near Station
149. Near the match box (4:5 inches in length) there are eroded blocks of an arenite bed which
are in process of being contributed to the nearby littoral sediments. Field evidence along the
coast between Aberystwyth and Borth suggests that a considerable amount of such erosion takes
place there.

Fic. 2. Boulder beach in the vicinity of Station 131, about 2 km. north of Sarn Bwch.
Boulders have been eroded from the glacial debris here and have formed a natural rip-rap. Never-
theless, considerable quantities of sand are being eroded from the cliffs and washed past the
boulders into the bay. The view is looking north. This region and the coastal section at
Towyn contribute much detritus to the bay.

Pee ASIE Ry 2

Bull. Br. Mus. nat. Hist. (Miner.) 2, 2

PLATE 3

Fic. 1. Thin section photomicrograph of sample from Station 162 (BM 1964, 227) near town
of Machynlleth. Note that all grains are lithic fragments, primarily of slate. River deposits
in the upper Afon Dyfi are all similar to this example in composition, although some local
variations in texture are observed along the banks. This and subsequent Afon Dyfi samples
are from sites of active transport. (Crossed nicols, x 30.)

Fic. 2. Thin section photomicrograph of sample from Station 161 (BM 1964, 226), down-
stream from Machynlleth. While the sediment is composed of lithic fragments resembling
those at Station 162 (PI. 3, fig. 1), there is a noticeable decrease in grain size. (Crossed nicols,
X 30.)

Bull. Br. Mus. nat. Hist. (Miner.) 2, 2 PA Ami iE 3

PLATE 4

Fic. 1. Thin section photomicrograph of sample from Station 159 (BM 1964, 224) in the
“lower river—upper estuary ’’ transition environment of sedimentation on the Afon Dyfi.
Note the pronounced appearance of fine quartz sand along with the larger lithic fragments.
This thin section shows the textural control of composition at this site. Since it is within tidal
range, the quartz sand is presumed to be related to the estuary. (Crossed nicols, x 30).

Fic. 2. Thin section photomicrograph of sample from Station 163 (BM 1964, 228) at the
head of the Dyfi estuary. Notice here that the predominant clastic is quartz sand and that
lithic fragments are much subordinate. This is a “ typical’’ estuary sand, quartzose and well
sorted. (Crossed nicols, x 30.)

Bull, Br. Mus. nat. Hist. (Miner.) 2, 2 PLATE 4

PLATE 5

Fic. 1. Thin section photomicrograph of sample from Station 148 (BM 1964, 212) near the
point where Sarn Wallog joins the coast. Note the mixed lithic (arenites and silty shales)
and quartz assemblage. This sample was collected at the base of the coastal cliffs of exposed
Aberystwyth grits. The abundance of large, angular arenite fragments is suggestive of their
proximity to source ; furthermore, such lithic grains are quickly reduced in size with transport
in a relatively high energy zone. (Crossed nicols, x 30.)

Fic. 2. Thin section photomicrograph of sample from Station 144 (BM 1964, 208), also from
the cliff-backed, littoral zone south of Borth. Notice here the similarity to Station 148 (Pl. 5,
fig. I) in grains present, 1.e., a mixture of quartz and lithic types. Some of the quartz suggests
affinities with vein quartz in the nearby outcrops. Regardless of local sorting, the compositional
suite remains much the same as shown here, with, perhaps, a slight increase in quartz grains
in the finer sediments. (Crossed nicols, x 30.)

Bull. Br. Mus. nat. Hist. (Miner.) 2, 2 IPAL IAD.

PLATE (6

Fic. 1. Thin section photomicrograph of sample from Station 150 (BM 10964, 214)
collected on the south end of the beach at Borth. This sample shows the mixture of both lithic
and quartz varieties in the littoral zone, an influence of local source and littoral sorting. Slate
fragments in this sample resemble slate outcropping nearby; furthermore, the slate/arenite
ratio of the Aberystwyth beds increases towards the north, as determined from measurements
along the coast (Prof. Alan Wood, personal communication). (Crossed nicols, x 30.)

Fic. 2. Thin section photomicrograph of sample from Station 7 (BM 1964, 91), a well-sorted,
sub-lithic sand offshore from Borth. Although this station is only 1-5 km. offshore, the index
of sorting (std dev.: 0-15; So: 1-06) is pronounced. Lithic (arenite) fragments constitute
slightly over 8% of the grains counted, as against 10:0% or over at Station 150 (Pl. 6, fig. 1) on
the coast. (Crossed nicols, x 30.)

Bull. Br. Mus. nat. Hist. (Miner.) 2, 2 PP AvhEn6

.
PRINTED IN GREAT BRITAIN
BY ADLARD & SON LIMITED
BARTHOLOMEW PRESS, DORKING