26. TEXTURAL AND COMPOSITIONAL VARIATIONS IN DOLERITE UNITSFROM HOLE 395A

G. Propach, S. Lee, and E. Prosser, Mineralogisch-Petrographisches Institut der Universitàt München,Theresienstr. 41, D-8000 München 2, F.R.G.

INTRODUCTION

The section of tholeiitic dolerite recovered fromHole 395A is of special interest because it is an almostcontinuously cored formation, 16 meters thick, whosecontact relationships are, for the most part, well pre-served.

The aim of this study was to examine the relation-ship between the dolerites and surrounding rocks andto determine to what extent, and in which respects,these dolerites can be related to continental occur-rences. Specific aspects discussed here are composi-tional and textural variations, differentiation effects,and cooling behavior.

The location of samples and stratigraphic relation-ships in this segment of Hole 395 A are given in Figures1 and 2.

PETROGRAPHY

Contact Relationships

The dolerite section discussed here occurs at about610 meters sub-bottom and is bounded above andbelow by the aphyric basalts of core Sections 61-1 and64-2. This segment (see Figure 1) is made up of upperand lower dolerite units having cored lengths of 4.2and 10.4 meters, respectively. These two units areseparated by a layer of fine-grained, aphyric basalt, 50cm thick.

Fine-grained marginal dolerite facies were recov-ered from three of the four dolerite-aphyric basaltcontacts. The fourth contact, the base of the upperdolerite unit, was not recovered. Only one of the firstthree contacts mentioned was continuously cored andfully recovered. This contact is represented by the basalchilled margin of the lower dolerite unit (Sample 64-2,129-132 cm), basaltic glass fragments (Sample 64-2,133 cm), and aphyric basalt. The possibility that thebasalt glass might belong to the chilled margin of thedolerite {Initial Core Descriptions) is slight, since thechemical composition of this glass (microprobe data,Melson, this volume) resembles more closely that ofthe underlying aphyric basalts than that of the dolerite.

Mineral Constituents

The minerals present in the dolerite can be classifiedaccording to their order of crystallization. The succes-sive stages can be summarized as follows:

First stage, Phenocrysts:Plagioclase, olivine, clinopyroxene, chromite.Main stage, Groundmass:

Plagioclase, clinopyroxene, olivine, titanomagnetite,Mesostasis:1

Sodic Plagioclase, clinopyroxene, titanomagnetite,ümenite, apatite,sulfide globules, glass.Postmagmatic stage, Alteration minerals:Brown, green, pale green hornblende, phillipsite,saponite, chlorite, low quartz, goethite, calcite.

Grain Size: Phenocrysts

Sinking or floating behavior of the phenocrysts andthe corresponding changes in grain size cannot bereliably determined through mean diameter measure-ments, because a large portion of the smaller pheno-crysts is often camouflaged by the groundmass compo-nents of the dolerites. Thus, only the maximum diame-ters of the individual phenocryst minerals in individualsamples (thin sections and polished rock slabs) werechosen to demonstrate grain size changes. Plots ofthese measurements along the profile produced veryerratic curves. These were smoothed by integratingmeasurements of two or three adjacent samples.

As can be seen in Figure la, there is no significantvariation in the maximum diameters of the clinopyrox-ene phenocrysts. The same holds true for olivine, withthe possible exception of the upper margin of the lowerdolerite unit, where depletion of the larger phenocrystsmay have occurred. Also noteworthy in this respect isthat the olivine phenocryst diameters are largest in theoutermost margins and decrease sharply in the sub-marginal zones. Such distribution behavior contrastswith that of continental occurrences and experimentalresults (Bhattacharji, 1967). Phenocrysts would beexpected to concentrate in the median portions of theflowing magma.

The Plagioclase phenocrysts are larger in the me-dian zones of the dolerite units than in the margins.The less calcic outer zones of these phenocrysts are lessthan 0.15 mm in width, so that post-intrusive growthcannot have contributed to any great extent to the totallength of the phenocrysts. This accumulation of largerphenocrysts could have resulted from the effects of flowdifferentiation (Bhattacharji, 1967).

Grain Size: GroundmassThe main trend of grain size variation in the doler-

ites—chilled margins giving way to coarser grained

1 The mesostasis is here defined as the very fine grained intersti-tial material between the groundmass crystals.

529

G. PROPACH, S. LEE, E. PROSSER

Section Interval (cm)

61-1

61-2

61-3

61, CC

62-1

63-1

63-2

63-3

63-4

63, CC

64-1

64-2

140-150-

75- 85-

90-110-

70- 80-

0- 10-

135-142-

80- 90-

17- 20H

75- 84-

17- 22-

115-122-

134-138-

59- 6 1 -

137-142-43- 45-77- 82-

112-117r120-132

0 0.2 mm 0 2 4µmj I I

.106 o"122 o

Figure 1. (A) Maximum phenocryst lengths for clinopyroxene, olivine and Plagioclase. Data from two to three sampleshave been combined to obtain more representative figures. Small humps in the curves are meaningless. (B) Mean grainsize of groundmass clinopyroxenes. (C) Maximum diameters of apatite needles in the mesostasis Shaded areas representphyric basalt units. Munich sample numbers are listed on the left of the second vertical line, DSDP and British Museumsamples are represented by crosses and circles, respectively, and are to the right.

interior portions—is described in Chapter 7 (thisvolume). A more detailed analysis was achieved bymeasuring the mean diameters of the groundmassclinopyroxenes.

Clinopyroxene proved most suitable for this pur-pose, since it is fresher than the olivines, more regularin shape than the plagioclases and titanomagnetites,forms few microphenocrysts, and is ubiquitousthroughout the dolerite.

The results of these measurements are plotted inFigure lb. The curves are roughly symmetric, but lossof core and the resultant shortening of true sectiondistorts the m. The basal 3 meters of the lower doleriteunit were recovered continuously, and the shape of thecorresponding curve would seem to be representativeof the true crystallization trend, which in turn wouldreflect the course of the cooling trend.

Modal Composition

Dolerite samples adjacent to the uppermost andlowest contacts (Table 1, No. 1 and 2) show strongenrichment in phenocryst content. This cannot be theresult of simple floating or sinking behavior, since both

Plagioclase and olivine are enriched in the basal aswell as roof margins. The amount of olivine that mayhave been added to the basal margin through crystalsettling is small: the rate of increase in phenocrystcontent from sub-marginal to marginal zones is almostequal for both the tops and bottoms of the doleritesills.

Samples from the sub-marginal zones of the upper-most and lowest contacts (Table 1, No. 4 and 5), aswell as samples from the uppermost recovered part ofthe lower unit (Table 1, No. 3) are almost identical incomposition. The proportion of phenocryst content inthe sub-marginal zones is similar to that of the porphy-ritic basalts (Table 1, No. 6) which, according to theInitial Core Descriptions, are the chemical equivalentsof the dolerites.

The lack of any noticeable selective depletion ofphenocrysts in the upper part or accumulation of thesame in the lower part of the dolerites indicates rapidsolidification, to such an extent that crystal movementwas inhibited. Phenocrysts of olivine or clinopyroxeneof the size found in the dolerites (up to 5 mm, seeFigure la) would sink with a velocity of a few centi-

530

TEXTURAL AND COMPOSITIONAL VARIATIONS IN DOLERITE UNITS

F

X 30 40

Figure 2. A-F-M plot of sample groups from Hole 395A. a = aphyric basalts, b = phyric basalts, c = dolerites from core sec-tions 61-1 to 64-2. I = late chilled dolerite, Palisades sill, 2 = chilled basal dolerite, Palisades sill, 3 = average tholeiite,Prinz et al., 1976., 4 = average aphyric basalt (395A), 5 = average dolerite, Palisades, 6 = average phyric basalt (395A),7 average dolerite (395A), 8 = lower chilled margin, lower dolerite (395A) Solid and dashed lines represent the Skaer-gaard and Palisades differentiation trends, respectively.

ters per hour (Shaw, 1965). The accumulation ofnoticeable amounts of olivine could thus take placewithin a few hours. Calculations on the basis of datagiven by Jaeger (1968) indicate that even such a shortinterval was not available in the margins. The drop intemperature from 1150° to 1100°C (where crystalsinking is assumed to cease because of lack of freespace) occurs over an interval of 35 to 0.6 hours,respectively, at corresponding distances from the con-tacts of 1 meter and 0.1 meter. The presumably highwater content of the surrounding rocks may haveincreased cooling rates along the immediate contacts.Rapid cooling is further indicated by the persistence ofthe undevitrified glass from the subjacent aphyricbasalts in immediate contact with the overlying chilledmargin of the dolerites.

MesostasisThe mean amount of mesostasis occurring in nine

dolerite samples from the lower dolerite unit is given inTable 1, No. 7. Individual plots of these samples areincluded in Figure 3. Little systematic variation inmesostasis content could be observed within the

coarser grained median portion of the dolerite. Theminimum shown in the lower unit results from theformation of secondary minerals, which is most exten-sive in this zone and preferentially occurs in themesostasis.

The minerals in the mesostasis are Plagioclase (Annear 30%, high-temperature properties), acicular clino-pyroxene, apatite needles, titanomagnetite, ilmenite,and sulfide globules. A felty aggregate, pale yellow toolive-green, presumably results from devitrification ofa residual glass. Clinopyroxenes in the groundmassbordering the mesostasis generally have light brownmargins. In one instance, (Sample 63-4, 134-138 cm),the clinopyroxenes show green rims.

Migration of the residual melt, now represented bythe apatite-bearing mesostasis, could have causedvariations in the phosphate content of the dolerites.Nothing of this kind was found. The steady amounts ofphosphorus (see Table 2) merely reflect the modalproportion of the mesostasis in the dolerites. Themaximum of the mesostasis curve is tentatively inter-preted as being the result of differing cooling rates.Such differences are not visible in the plot of sizes of

531

G. PROPACH, S. LEE, E. PROSSER

TABLE 1Average Modal Compositions of Dolerites and

Phyric Basalts From Hole 395A (in vol. %)

Phenocrysts

TABLE 2Dolerite Compositions (per cent) Hole 395A

1234567

Plag

30.024.818.119.016.724.160.2 ± 2.3

Ol

6.68.23.63.74.85.27.5 ± 1.7

Cpx

0.50.50.71.30.71.0

23.2 ± 2.4

Cms

62.966.577.676.077.969.7

2.3 ± 0.9 5.9 ± 1.4 0.5 ± 0.4

Note: 1. Upper margin, upper dolerite unit, mean of Samples 61-1, 122 cm and 118-146 cm. 2. Lower margin, lower dolerite mean of Samples 64-2, 122 cm and131cm. 3. Upper margin, lower dolerite, mean of Samples 62-1, 72cm and 82cm.4. Sub-marginal zone, upper part of upper dolerite unit, mean of Samples 61-1,140cm and 61-2, 21-37cm. 5. Sub-marginal zone, lower part of lower unit, meanof Samples 64-2, 106cm and 112cm. 6. Porphyritic basalt (P4), mean of fivesamples. 7. Mean and standard deviation of nine samples from the lower doleriteunit. Plag. = Plagioclase, Ol = olivine, Cpx = clinopryoxene, Gms = groundmass,Ore = Opaque minerals, Mss = mesostasis, Sec = secondary minerals

61 1

61-2

613

6 1 , CC

62-1

63 1

63-2

63-3

63-4

63,-CC"

64-1

64-2

3 5

f»

10% ,, , , I l

•

•

100% ol-alter.

\

\

\

\

I

9<t c :t cc g<>e p h h DI qz

Figure 3. Volume per cent of alteration of olivine (esti-mated), and presence of secondary minerals. The amountof mesostasis in Sections 63-4 and 63, CC is lowered bysecondary mineralization, cc = calcite, cct = colorlesschlorite, get = green chlorite, goe= goethite, hbl= horn-blende, ph =phillippsite, qz = quartz. Saponite is presentin all samples.

groundmass clinopyroxene (Figure lb) and diametersof acicular clinopyroxene in the mesostasis (not plot-ted).

Alteration

The effects of alteration are most prominent in azone 3 to 4 meters above the lower contact of thelower sill. Secondary minerals in this zone constitute upto 40 per cent of the total volume. The qualitativemineralogical composition of the secondary mineralassemblages are shown in Figure 3.

Alteration is indicated by the growth of brownhornblende (up to 0.5 mm in width) on clinopyroxene.The brown hornblende in turn shows terminations ofgreen and pale green hornblende. The total amount ofhornblende is less than 0.5 per cent. Phillipsite, mostlyas aggregates of radiating fibers, can be found in vugs,

61-1,140-150

61-2,75-85

Sample (Interval in cm)

61-3 62-1,90-110 70-80

63-1,0-10

63-1,135-142

SiO2

TiO 2

A12O3

F e 2 O 3

FeOMnOMgOCaONa2OK2OP2O5H 2 O +

H 2 O "

48.631.18

16.443.315.230.158.04

11.792.680.160.121.480.80

49.061.11

16.962.016.360.138.13

11.782.550.100.111.210.50

Trace elements ^µµm;

Rb 9 6SrZr

12765

13565

48.691.13

16.612.016.500.148.44

11.982.350.090.121.490.44

5118113

49.591.21

16.303.535.070.157.90

11.922.390.150.081.120.59

8126104

49.331.14

16.951.976.200.148.28

11.652.380.090.111.250.51

412581

49.041.05

16.723.954.620.148.55

11.852.180.140.111.070.93

6120118

63-2,80-90

63-375-84

Sample (Interval in cm)

634, 64-1,115-122 137-142

64-2,77-82

64-2,112-117

SiO2

TiO 2

Al 2 θ3

FeoMnOMgOCaONa2OK 2OP2O5H 2 O +

H 2 O-

49.161.14

16.723.105.160.158.14

12.122.410.130.111.130.53

47.571.10

17.963.434.670.157.32

12.392.700.120.121.730.75

Trace elements (ppm)

RbSrZr

9128148

7124109

48.040.87

16.863.514.000.149.09

12.192.290.090.111.790.96

7118131

49.161.04

17.452.495.300.137.90

12.132.480.110.121.240.40

9121105

47.711.10

17.173.295.250.148.30

12.382.600.110.121.380.71

14123102

45.691.08

15.403.285.020.148.42

11.452.570.140.111.130.47

12996

SiO2

TiO 2

Al 2 θ3F e 2 O 3

FeOMnOMgOCaONa2OK2OP2O5H 2O+

H2O-

395A-64-2,

120-132

48.471.00

16.933.254.780.148.94

11.922.300.140.111.400.60

Sample (Interval in cm)

395-17-1,

56-69

47.450.35

16.731.154.950.11

11.628.993.610.210.024.480.33

r1Trace elements (ppm)

Rb 8SrZr

122105

31388

395-18-1,

61-70

40.570.010.745.555.170.11

40.270.840.020.020.016.380.30

395A-15-5,0-11

48.561.12

19.432.045.630.116.98

12.032.480.090.141.020.38

Calibration standards

<l342

< l14492

395A-63-1,

108-116

49.531.08

17.531.975.770.138.13

12.072.140;090.091.320.16

1_l

712387

altered olivines, and veins. Very rare euhedral low-quartz crystals occur in vugs filled with saponite.

Alteration of olivine is most prominent in thecoarse-grained parts, but exceptions do occur. Minerals

532

TEXTURAL AND COMPOSITIONAL VARIATIONS IN DOLERITE UNITS

replacing olivine are saponite, phillipsite, calcite, "li-monite," and goethite. There is a weak positive corre-lation between the degree of alteration, H2O+ content,and Fe2O3/FeO ratios (see Figure 4).

CHEMISTRY

MethodsAll analyses were made on an X-ray spectrometer,

sample presentation being in the form of pressed glasspowder tablets. Primary standards USGS W-l, ZGIB.M., CSRM MRG-1, USGS PCC-1, NIM N, NIM D.,as well as secondary in-house standards, were used asreference samples. Measurements on shipboard calibra-tion standards 395A-15-5, 0-11 cm, 395A-63-1, 108-116 cm, 395-17-1, 56-59 cm, and 395-18-1, 61-70 cm areincluded in Table 2.

Na2O and K2O values as well as Rb contents of thedolerites were determined on a flame photometercoupled to a lock-in amplifier (Cammann, 1972). FeOwas determined using a Teflon digestion vessel undernitrogen atmosphere. The digested sample was titratedwith potassium dichromate, the end point being deter-mined with an electrode coupled to an automatic

titrator/potentiograph. H2O (total) was extracted at1600°C with an induction furnace and carried over toa Karl-Fischer titrator with nitrogen.

Analytical ResultsIndividual dolerite analyses are shown in Table 2.

Average compositions of the dolerites, phyric andaphyric basalts, as well as for the basal chilled marginof the lower dolerite unit, are included in Table 3. Thephyric basalt average includes geochemical unitsP2,P3,P4 and P5. Compositionally, the dolerites can bereadily distinguished from the aphyric basalts throughtheir lower TiO2, Fe2O3, FeO, Na2O, and K2O, andhigher A12O3, MgO, and CaO contents. The differencesbetween the dolerites and the phyric basalts are gener-ally limited to the lower A12O3 and MgO contents ofthe latter.

As is shown in the A-F-M plot (Figure 2), the basalchilled margin (8), average dolerite (7), phyric (6),and aphyric (4) basalt compositions from Hole 395Aare arranged, in the same order of succession, roughlyparallel to, and in the direction of, the usual tholeiiticdifferentiation trends. The compositional fields of theaphyric (a) and phyric (b) units overlap slightly,

Section Interval (cm)

Fe2O3/FeO

.2 .4 .6 .8 1.0 1.0

%H20+

1.4 1.8 4

D.I.

20 25

61-1

61-2

61-3

61, CC

62-1

63-1

63-2

63-3

63-4

63, CC

64-1

64-2

140-150-

75- 85

90-110

70- 80-

0- 10-

135-142

80- 90-

75- 84

115-122-

137-142-

77- 82-112-117-129-132

-80- 87

-42- 47 X

-108-114 ©

86- 93 x

.29- 37 x•45- 54 ©

Figure 4. Variation ofFe2O3/FeO, H2O+, normative olivine and quartz and the differentiation index through two sills. Sam-

ple numbers to the left are from Munich collection, those on the right are from Bougault (w), Rhodes (×), and Zolotarev(O), this volume.

533

G. PROPACH, S. LEE, E. PROSSER

TABLE 3Average Compositions (in per cent) of Dolerite and Phyricand Aphyric Basalts, Compared to Chilled Basal Dolerite

(Sample 64-2,129-132 cm)

Rock Type

No. of Analyses

SiO2

TiO2

A12O3

Fe 2 θ3FeOMnOMgOCaONa2OK 2 OP2O5H 2 O +

H 2 O-FeO Tot.

F e 2 O 3

FeO

Mg

Mg+Fe

ChüledMargin

Dolerite

1

48.391.00

16.903.254.780.148.92

11.902.300.140.121.400.607.70

0.68

0.67

Dolerite

12

48.601.10

16.832.995.290.148.16

12.002.470.120.101.310.627.99

0.56

0.64

PhyricBasalt

22

48.821.20

17.562.905.380.147.29

12.242.500.14

0.970.537.98

0.54

0.62

AphyricBasalt

24

48.511.66

15.123.796.200.177.67

10.612.800.23

1.140.849.61

0 6 1

0.59

whereas the dolerites (c) are to a large extent enclosedin the compositional field of the phyric basalts, thusdemonstrating the chemical affinity of the latter tworock types.

The molecular Mg/Mg + Fetotal ratios for the chilledmargin, average dolerites, and phyric and aphyricbasalts follow much the same pattern as depicted inthe A-F-M diagram. Corresponding values are 0.67,0.64, 0.62, and 0.59, respectively. The basal chilledmargin thus represents the most primitive of theseunits.

Distribution trends of the major oxides and norma-tive minerals (and modes) in the upper and lowerdolerite sills are generally erratic (see Figure 4). This isreflected in the behavior of the Differentiation Index(D.I.) and normative olivine (and quartz) trends (Fig-ure 4). The olivine-quartz trend, although exhibiting a

net increase in normative olivine towards the basalchilled margin, shows variations of up to about 8 percent between the neighboring samples. The increase innormative olivine may be a consequence of crystalsettling effects too weak to have produced an olivine-rich layer but suggesting enrichment tendencies. Clino-pyroxene crystals are too scarce to have produced, orpermit detection of, any enrichment effects. Pro-nounced compositional variations in Sections 63-3 and63-4 (see Figure 4) correspond to increases in second-ary mineral content and to the presence of brecciatedzones in this part of the lower sill.

SummaryEffects of differentiation (crystal sinking or floating,

migration of residual or immiscible melts) are verylimited (sinking of olivines) or even undemonstrable(sinking of clinopyroxene, floating of Plagioclase, mi-gration of residual or immiscible melts). The effects ofdiffering cooling rates are, however, clearly developed:gradually increasing grain size in the groundmass,segregation of a (immiscible?) melt in the medianportions of the two units, now represented by themesostasis, and a more extensive hydrothermal miner-alization in the middle of the lower dolerite unit.

REFERENCES

Bhattacharji, S., 1967. Mechanics of flow differentiation inultramafic and mafic sills, J.Geol, v. 75, p. 101-102.

Cammann, K., 1972., Aufbau eines extrem empfindlichenFlammenfotometers für Analysen im Picogramm-Bereichmit Hilfe eines "lock-in" Verstàrkers, Messtechnik, v. 9, p.272-277.

Jaeger, J. C, 1968. Cooling and solidification of igneousrocks. In Hess, H. H. and Poldervaart, A. (Eds.), Basalts,v. 2: New York (J. Wüey and Sons), p. 503-536.

Prinz, M, Keil, K., Green, J. A., Reid, A. M., Bonatti, E., andHonnorez, J., 1976. Ultramafic and mafic dredge samplesfrom the equatorial Mid-Atlantic Ridge and fracturezones, J. Geophys. Res., v. 81, no. 23, p. 4087-4103.

Shaw, H. R., 1965. Comments on viscosity, crystal settlingand convection in granitic magmas, Am. J. Sci., v. 263, p.120-152.

534