24. rock magnetic properties, magnetic mineralogy, …
TRANSCRIPT
Mével, C , Gillis, K.M., Allan, J.F., and Meyer, P.S. (Eds.), 1996Proceedings of the Ocean Drilling Program, Scientific Results, Vol. 147
24. ROCK MAGNETIC PROPERTIES, MAGNETIC MINERALOGY, AND PALEOMAGNETISMOF PERIDOTITES FROM SITE 895, HESS DEEP1
Paul R. Kelso,2 Carl Richter,3 and Janet E. Pariso4
ABSTRACT
Serpentinized oceanic peridotites collected during Ocean Drilling Program Leg 147 at Site 895 in the Hess Deep were stud-ied magnetically to further our understanding of the magnetization of the oceanic lithosphere. The majority of the samples aredunites and harzburgites, with varying degrees of serpentinization. Rock magnetic studies suggest that the dominant magneticmineral is relatively pure pseudo-single-domain magnetite. Optical observations reveal that the magnetite has a secondary ori-gin related to the serpentinization process. The dunites and the harzburgites have median natural remanent magnetization val-ues of 3.2 and 1.2 A/m and magnetic susceptibility values of 0.053 and 0.023 (S1 volume units), respectively. The NRM is thedominant component of magnetization and has a magnitude similar to values observed for oceanic basalts of Layer 2a. Simi-larly serpentinized peridotites residing within the oceanic lithosphere are likely contributors to magnetic anomalies. Thermaland alternating-field demagnetizations usually yield a stable remanent magnetization direction that often has an inclinationmuch greater than expected for the time-averaged geomagnetic field (>5°) at Site 895. This disparity suggests that the samplesacquired their magnetization and were subsequently reoriented. Hole 895E samples have relatively coherent inclinations (30°)that become shallower in the bottom 30 m of the hole. Inclinations in Hole 895D vary from +60° to -60°, with no systematictrends evident with depth in the hole, suggesting differential vertical rotation of blocks on a scale of a few meters. Experimentalevidence suggests that room-temperature magnetic viscosity does not significantly effect the magnetization intensity or direc-tion of these samples.
INTRODUCTION
Ocean Drilling Program (ODP) Leg 147 (Gillis, Mével, Allan, etal., 1993) collected hard-rock samples from the intrarift ridge of HessDeep in the central east Pacific Ocean (Fig. 1). The samples recov-ered at Site 894 were mostly gabbroic rocks from the lower oceaniccrust, whereas samples from Site 895 were dominated by serpenti-nized peridotites likely from the uppermost mantle (Gillis, Mével,Allan, et al., 1993). The magnetic properties of the samples of Site894 are discussed in a companion paper by Pariso et al. (this volume).This paper concentrates on the magnetic properties of samples fromSite 895, particularly from Holes 895D and 895E, which had thegreatest recovery and deepest seafloor penetration (93.7 m and 87.6m, respectively). These cores contained peridotites from the oceaniclithosphere, with harzburgites dominant in Hole 895D and dunites inHole 895E. This study more than doubles the number of serpenti-nized peridotite samples analyzed magnetically from ocean drillingstudies (Dunlop and Prévot, 1982; Smith and Banerjee, 1985; Bina etal., 1990; Bina and Henry, 1990; Hamanoetal, 1990). Previous stud-ies involved sites where peridotite samples were found at "anoma-lously" shallow depths in the oceanic lithosphere, for example, infracture zones and complicated tectonic regions. Thus, the process offormation and subsequent alteration may not be representative of"typical" oceanic crust, and the resulting magnetic properties fromthese anomalous sites may not be representative of the oceanic crust
'Mével, C, Gillis, K.M., Allan, J.F., and Meyer, P.S. (Eds.), 1996. Proc. ODP, Sci.Results, 147: College Station, TX (Ocean Drilling Program).
2Lake Superior State University, Sault Ste. Marie, MI 49783, U.S.A.pkelso @ lakers.lssu.edu
3 Ocean Drilling Program, 1000 Discovery Drive, Texas A&M University ResearchPark, College Station, TX 77845, U.S.A. carl_richter@ odp.tamu.edu
4School of Oceanography, University of Washington, Seattle, WA 98195, [email protected]
200 kmmm Cocos-Nazca
••• • - ^ c r u s t
^ × fault
—2°30'NCocos
«- «*
Figure 1. Location map of Sites 894 and 895 and the generalized tectonic set-ting of Hess Deep near the junction of the East Pacific Rise (EPR) and theGalapagos spreading center (after Lonsdale, 1988).
in general. The Site 895 peridotites formed at a typical fast-spreadingcenter, the East Pacific Rise, and are now exposed at Hess Deep, atthe tip of the westward-propagating Cocos-Nazca Ridge. A review ofthe geologic setting and tectonic history at Site 895 can be found inarticles by Lonsdale (1988), Francheteau et al. (1990), Gillis, Mével,Allan, et al. (1993), MacLeod et al. (this volume), and Mével and Gil-lis (this volume).
Samples from Hole 895E were more altered than those of Hole895D, with the former often containing >90% secondary minerals.All peridotite samples were highly serpentinized, typically 70% ormore. For a detailed analysis of the igneous and metamorphic petrol-ogy of Site 895, see Arai and Matsukage (this volume), Früh-Greenet al. (this volume), and Mével and Stamoudi (this volume).
405
P.R. KELSO, C. RICHTER, J.E. PARISO
A continuing debate among geophysicists is the relative contribu-tions of the different oceanic layers to marine magnetic anomalies(e.g., Toft and Arkani-Hamed, 1992,1993). A major limitation to ourunderstanding of marine magnetic anomalies has been our lack ofknowledge of the magnetic properties of units below the extrusive ba-salts of Layer 2a, largely due to the scarcity of samples from the deep-er units. The peridotites of Site 895 provide a unique new data set ofthe magnetic properties for specimens from near the crust/mantleboundary.
EXPERIMENTAL PROCEDURES
Measurements of saturation magnetization (Js) vs. temperature(T) were made from room temperature (22°C) to 700°C on a vibratingsample magnetometer equipped with a furnace, at the Institute forRock Magnetism (IRM) at the University of Minnesota. The temper-ature was monitored by a thermocouple that was in contact with thesample ( IOO mg). The sample was wrapped in silver foil to facilitatea homogeneous temperature distribution. The applied magnetic fieldwas 500 mT. The sample chamber was evacuated and then backfilledwith helium several times before each Js vs. Tran to purge the systemof oxygen, to reduce the chance of oxidation, and to facilitate heattransfer. Curie temperatures (Tc) were calculated graphically fromthese curves. Hysteresis loops confirm that 500 mT is sufficient tomagnetically saturate the samples. Hysteresis loops were generated atthe IRM for a variety of samples whose masses ranged from 0.01 to10 g. A Quantum Design superconducting susceptometer was used tomeasure the temperature dependence of saturation remanence as itwas heated in zero magnetic field. Samples ( IOO mg) were saturatedby a 500-mT field at 15 K (SIRM[15K]). Measurements were takenevery 5 to 10 K as the sample was heated from 15 K to 300 K in zeromagnetic field (i.e., less than a few millitesla).
The samples used for directional analysis were cylindrical speci-mens drilled from the working half of the core. The minicores wereapproximately 2.5 cm in diameter and 2.2 cm in length. NRM mea-surements were made with a 2G magnetometer aboard the JOIDESResolution and at the IRM, and with a CTF cryogenic magnetometerat Texas A&M University. The NRM was stepwise demagnetizedwith either alternating field (AF, maximum field 100 mT) or thermal-ly in zero field (maximum temperature 600°C). For long minicores,the 2.5- by 2.2-cm trimmed minicore was AF demagnetized and theremaining end piece was thermally demagnetized. Thermal demag-netization was conducted in a Schonstedt thermal demagnetizer inwhich the sample was held at temperature for 30 to 45 min dependingon sample size. AF demagnetization was accomplished using a sin-gle-axis Schonstedt AF demagnetizer.
Magnetic Mineralogy and Rock Magnetic Properties
An understanding of marine magnetic anomalies and the magnet-ic properties of the oceanic lithosphere requires knowledge of theconcentration and composition of the dominant magnetic minerals inrepresentative rocks. The mineralogy of the phases that contribute tothe magnetization of samples from Site 895 was examined throughmagnetic studies, including measurements of saturation magnetiza-tion vs. T, saturation remanent magnetization vs. T, susceptibility vs.T, hysteresis loops, and AF and thermal demagnetization of theNRM.
The measured peridotite samples have Curie temperatures in thenarrow temperature range between 560° and 585°C, with a medianvalue of 578°C (Fig. 2; Table 1). The curves are essentially revers-ible, implying that little chemical change occurred during the heatingprocess. This suggests that the dominant magnetic mineral is nearlypure magnetite (Tc = 585°C). Minor substitution of other elements(e.g., Ti, Al, Mg) decreases the Curie temperature of magnetite.There is no evidence of the phase transition that is commonly ob-
6000
o 5000
c I 4000| "o i 3000
c 2000 -
1000-
Section147-895E-1R-2,105 cm •
Section 147-895E-3R-2,136 cm j
Section 147-895E-5R-2,048 cm :
i i i nX~ in100 200 300 400 500
Temperature (°C)600 700
Figure 2. Representative saturation magnetization vs. temperature curves thatwere used for Curie temperature determinations.
served for oxidized titanomagnetites characteristic of oceanic basalts(e.g., Prévotetal., 1981; Beske-Diehl, 1990).
Heating the SIRM[15K] of a few typical samples in zero magneticfield (Fig. 3) reveals a large decrease in magnetization between 100and 120 K due to a lattice transition (Verwey) in magnetite. There isno evidence of the hematite transition that occurs at around 260 K.The change in slope of the curve between 30 and 40 K has several po-tential causes: (1) it could be due to the thermal demagnetization ofextremely fine-grained magnetites, (2) a lattice transition in pyrrho-tite that occurs at 35 K, (3) the magnetic ordering of an ilmenitephase, or possibly (4) a magnetic ordering transition in an iron silicatephase. If superparamagnetic (SP) magnetite was abundant therewould be a dramatic decrease in the magnetization at low temperature(15-80 K) and this may continue even at higher temperature (150-250 K), but no such decrease was observed. Thus, it is concluded thatsamples do not contain a significant volume of magnetite that is SPat room temperature.
During thermal demagnetization the natural remanent magnetiza-tion decreases to zero at between 550° and 600°C (Fig. 4). There areno other changes in slope of the demagnetization curve, which sug-gests that no other minerals, such as titanomagnetite or pyrrhotite,contribute significantly to the NRM.
Hysteresis loops were analyzed to determine the bulk magneticcoercivity, magnetic grain size, and concentration of magnetic min-eral. The values of saturation magnetization (Js, extrapolated to zerofield), saturation remanent magnetization (/„), coercivity (Hc), andcoercivity of remanence (Hcr) are recorded in Table 1. A Day et al.(1977) plot of the hysteresis parameters (Fig. 5A) shows that the peri-dotites generally plot in the lower portion of the pseudo-single do-main (PSD) region whereas the gabbronorites of Site 894 plot in thePSD-multidomain (MD) boundary region (Pariso et al., this volume).The harzburgites have on average higher JJJS and lower HJHC ratiosthan the dunites (Fig. 5B). This suggests that on average the effectivemagnetic grain size of the harzburgites is slightly less than that of thedunites.
These magnetic measurements lead to the conclusion that PSDmagnetite is the dominant magnetic mineral in the serpentinized dun-ite and harzburgite samples of Site 895. This is consistent with previ-ous studies of serpentinized oceanic peridotites (e.g., Dunlop andPrévot, 1982; Krammer, 1990) and theoretical arguments (Frost,1985, 1991; Toft et al., 1990). Observations made by reflected lightmicroscopy show that the opaque minerals mainly occur within theserpentinized groundmass and along altered fractures within the lessaltered grains, particularly due to the serpentinization of olivine (Gil-lis, Mével, Allan, et al., 1993). This suggests that the opaque miner-
406
Table 1. Summary of the magnetic properties of samples from Site 895.
Samplenumber (cm)
147-895B-1R-1,541R-1,64
147-895C-3R-1.53R-1, 1393R-2, 33R-2, 194R-1,24R-1,234R-1, 344R-1, 834R-2, 1054R-3, 58
147-895D-2R-1, 252R-1.472R-2, 222R-2, 252R-2, 392R-2, 472R-2, 783R-1.303R-1.454R-1.414R-2, 254R-2, 394R-2, 1234R-3, 44R-3, 154R-3, 724R-3, 1104R-4, 194R-4, 754R-4, 844R-4, 1004R-5, 34R-5, 85R-1.525R-1, 1205R-2, 1006R-1.326R-1.756R-1.997R-1.657R-1, 847R-1, 1197R-2, 137R-2, 927R-2, 1078R-1.458R-2, 48R-2, 1199R-1.789R-1, 1209R-1, 1299R-1, 140
147-895E-1R-1,47lR-1,771R-2, 721R-2, 1051R-2, 1251R-2, 129
Depth(mbsf)
0.540.64
18.3019.6919.8119.9027.9328.1328.2428.7330.0331.06
16.2016.4717.6417.6317.8117.8918.2026.3026.4535.1135.8435.9636.8237.0337.1437.7138.0938.6439.2039.2939.4539.7639.8143.8244.5045.7455.3255.7553.2065.2567.5265.7966.2066.9967.1474.7575.7276.8684.7885.2085.2985.40
0.470.772.162.492.692.73
Density
(1000 kg/m3)
2.68
2.392.70
2.44
2.66
2.74
2.98
2.612.722.46
2.622.502.622.732.722.70
2.632.65
2.752.832.772.73
2.682.84
2.782.732.672.90
2.522.66
2.65
2.502.352.612.90
2.452.53
2.54
2.782.742.522.54
2.44
Susceptibility(S1)
0.00720.0074
0.028650.0233
0.04260
0.014720.05770.02370.02030.0008
0.006210.01030.03337
0.02610.03650.020770.006870.00580.003700.00318
0.014980.010630.01440.003350.00370.003350.003560.00350.007820.00480.00350.02480.004990.01150.00030.03350.031650.01993
0.02800.03130.03250.021480.02840.0003
0.045020.06730.06010.07087
0.00030.000240.07450.068510.09500.08062
J,(A/m)
0.190.19
0.750.61
1.12
0.391.520.620.530.021
0.160.270.88
0.680.960.550.180.150.100.08
0.390.280.380.0880.0970.0880.0940.0920.210.130.0930.650.130.300.0090.880.830.52
0.740.820.850.560.750.008
1.181.771.581.86
0.0080.0061.961.802.502.12
NRM(A/m)
0.630.47
0.992.03
2.07
1.222.742.842.110.015
0.471.022.16
0.811.350.73d.:s0.260.570.40
1.011.071.430.380.430.540.480.980.690.610.521.480.520.780.0113.251.922.45
0.392.012.040.963.210.12
2.4725.04
9.97
0.0100.0105.064.997.365.99
0
3.32.4
1.33.3
1.9
3.21.84.64.00.7
2.93.72.5
1.21.41.31.51.75.94.8
2.63.83.84.34.56.25.2
10.63.44.85.52.34.02.61.23.72.34.7
0.52.42.41.74.3
15.1
2.114.26.3
1.21.72.62.82.9
MDF(mT)
12
15
10.9
10
11
19.214
712.6
12.514.9
19.3
15
17.120
12
1513
17.814.4
12
1213.71019
13.618
44
1014.8
Λ(Am2/kg)
0.171
0.787
1.1402.3600.8240.867
0.797
0.377
1.2362.2070.903
0.140
0.167
0.2450.5590.805
0.114
0.113
0.264
0.100
0.9031.990
1.693
1.320
2.8072.510
1.6903.150
8.92E-41.58E-3
3.210
3.120
Jn
(Am2/kg)
0.039
0.291
0.2520.2600.1540.187
0.231
0.089
0.2950.6080.127
0.018
0.030
0.0380.0660.092
0.019
0.023
0.050
0.022
0.2580.447
0.287
0.226
0.2180.434
0.3110.273
2.26E-41.79E-4
0.446
0.305
Hc
(mT)
14.15
29.5
16.579.02
12.3517.78
19.22
17.30
18.9421.0310.02
11.14
13.94
15.4610.909.09
13.56
17.60
14.40
17.35
16.4117.74
13.90
12.46
7.7712.32
10.807.90
14.662.96
10.48
7.86
Hcr
(mT)
25.47
45.7
28.3020.8424.1031.42
30.31
27.30
34.6534.8319.29
23.40
26.42
31.2022.9319.98
27.23
32.23
26.20
31.55
26.2029.31
26.10
23.44
20.3021.34
20.7020.90
44.35
20.00
16.59
JJJr
0.23
0.37
0.220.1100.190.216
0.29
0.236
0.2390.280.14
0.131
0.178
0.160.1190.114
0.163
0.207
0.188
0.22
0.290.225
0.17
0.171
0.080.173
0.180.087
0.234
0.139
0.098
HJHC
1.8
1.5
1.72.32.01.8
1.6
1.6
1.81.71.9
2.1
1.9
2.02.12.2
2.0
1.8
1.8
1.8
1.61.7
1.9
1.9
2.61.7
1.92.6
3.0
1.9
2.1
Magnetite(%)
0.19
0.86
1.22.60.900.94
0.87
0.41
1.32.41.0
0.15
0.1 8
0.270.610.88
0.12
0.12
0.29
0.11
1.02.2
1.8
1.4
3.12.7
1.83.4
9.7E-41.7E-3
3.5
3.4
Tc ave.(°C)
570
582
1
585
570
579
585
584
580
581
579
Primary direction (°)
Inc.
249-63
-13
1571056
25-139
13410919
116136
2153
49
344265
2320
-89
163-152
17675
241
37274118184-41
9452
223225
-60
-74318317
Dec.
5439
46
514644
-4438
-31-34-10
- 2 3-45
-9-11
38
594124
-6
-36
6131
-15
5956505449
16-59
35-61
17
474333
Lithology
HarzburgiteHarzburgite
HarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteTroctoliteDuniteDiabase
HarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteHarzburgiteGabbronoriteHarzburgiteHarzburgiteTroctoliteTroctoliteHarzburgiteHarzburgiteHarzburgiteHarzburgiteDuniteGabbronoriteDuniteHarzburgiteDuniteDuniteDunite
Olivine gabbroOlivine gabbroDuniteDuniteDuniteDunite
Measuringlab
TAMUJR
IRMJRUWA, IRMIRMUWA, IRMIRMTAMUJRTAMUJR
IRMJRIRMUWA, IRMTAMUJRIRMIRMJRIRMIRMUWA, IRMIRMIRMTAMUIRMJRIRMIRMTAMUIRMJRTAMUJRIRMJRJRTAMUIRMIRMUWA, IRMJRTAMUJRIRMJRJRUWA, IRMIRMJRTAMUIRM
JRIRMJRIRMTAMUIRM
Q
n
>Oz-n"-aSβ
O
£XT.ÇΛCΛ
-
oo
Samplenumber (cm)
1R-3, 24lR-3,491R-3, 551R-3, 86lR-3,951R-3, 1222R-1, 852R-2, 182R-2, 562R-2, 1073R-1.903R-2, 243R-2, 553R-2, 1363R-3, 144R-1.924R-1, 1024R-2, 124R-2, 464R-2, 584R-3, 65R-1, 1125R-1, 1455R-2, 485R-2, 1155R-2, 1325R-3, 195R-3, 356R-1, 1316R-2, 216R-2, 676R-2, 856R-2, 1116R-3, 176R-3, 686R-3, 1016R-4, 586R-5, 116R-5, 406R-5, 666R-5, 746R-5, 956R-6, 277R-1.817R-1, 1137R-1, 1357R-2, 1027R-3, 287R-3, 397R-3, 607R-3, 867R-3, 1257R-3, 1457R-4, 367R-4, 598R-1.998R-1, 1138R-1, 1208R-2, 168R-2, 358R-2, 578R-2, 708R-2, 908R-2, 1008R-3, 2
Depth(mbsf)
3.103.353.403.723.794.08
20.4321.1421.5222.0330.5031.2131.5232.3332.6140.4240.5242.2541.4041.5242.4751.0250.3550.8851.5551.7252.0952.2559.9160.3160.7760.9561.2161.4861.9962.3263.0063.8664.1564.4164.4964.7065.2569.0169.3369.5570.7271.4871.5971.8072.0672.4572.6573.0673.2978.8979.0379.1079.5579.7479.9680.0980.2980.3980.82
Density(1000 kg/
m3)
2.55
2.53
2.75
2.472.35
2.532.602.542.372.53
2.87
2.542.602.53
2.54
2.58
2.59
2.562.492.54
2.552.57
2.592.48
2.662.62
2.49
2.542.48
Susceptibility(S1)
0.05990.11350.08020.074780.11700.0004
0.05530.07690.05990.032620.07870.11540.133400.12540.046220.05840.06600.05010.05190.00030.06100.06750.052410.07630.05500.04940.05000.04930.04120.034740.04480.02720.02330.02500.02550.03080.01560.04580.04650.047180.045770.06610.05300.06930.059340.04640.03840.03790.03560.040390.06150.02390.01270.012610.04570.045300.04780.07380.05910.05920.037720.03730.06580.0470
h(A/m)
1.572.982.111.963.070.012
1.452.021.570.862.073.033.503.291.211.531.731.311.360.0071.601.771.382.001.441.301.311.291.080.911.180.710.610.660.670.810.411.20.22
1.241.201.741.391.821.561.221.011.000.941.061.610.630.330.331.201.191.251.941.551.550.990.981.731.23
NRM(A/m)
3.825.343.243.454.480.001
3.0712.997.581.558.80
10.34
4.673.333.212.953.533.930.0013.044.353.454.782.802.80
3.382.422.151.712.304.531.111.932.102.272.065.251.882.223.412.275.284.582.791.642.102.121.943.191.481.041.122.982.493.063.502.413.721.972.482.003.53
Q
2.41.81.51.81.50.1
2.16.44.81.84.33.4
1.42.72.11.72.72.90.11.92.52.52.41.92.2
2.62.22.41.53.27.41.72.92.65.51.74.31.51.82.01.62.92.92.31.62.12.31.82.02.43.13.4
2.52.12.41.81.62.42.02.51.22.9
MDF(mT)
I08.8
10.64
10.6
13
10
98.27
10.3
11.79
15.47
11
10.7144.4
71210
712.716.410.513.510.39.3
1214.1
10
241115.82222.5
23.379.5
8.322
h(Am2/kg)
1.8403.94
2.719
2.530
1.3702.360
4.6504.2902.510
1.320
2.745
2.550
2.020
1.360
2.1000.855
1.36
1.610
4.4302.640
1.7202.4100.9990.624
1.8201.580
2.6501.860
2.995
Table 1 (continued).
Jn
(Am2/kg)
0.3140.56
0.359
1.7E-5
0.458
0.2080.345
0.7120.4420.274
0.249
0.528
0.386
0.256
0.216
0.2640.1340.232
0.186
0.7180.341
0.2280.2410.2210.116
0.2090.231
0.3630.251
0.198
(mT)
12.1510.17
10.37
0.29
10.81
11.118.71
9.208.938.60
12.49
16.20
12.30
11.47
11.21
11.5712.1812.31
8.79
12.3010.10
11.499.26
17.8315.82
9.6211.46
11.0410.80
7.03
(mT)
22.3620.44
19.90
41.50
21.06
21.3417.42
19.1019.4418.30
23.70
27.95
23.70
24.43
22.72
23.9723.723.2
18.69
22.6519.80
22.6319.5529.6026.40
24.0522.05
21.9721.00
21.60
JrJJr
0.170.142
0.132
0.05
0.18
0.1520.15
0.1530.100.109
0.19
0.19
0.151
0.13
0.16
0.130.160.17
0.116
0.160.129
0.1330.100.220.19
0.110.146
0.140.135
0.07
HJHC
1.82.0
1.9
143.1
1.9
1.92.0
2.12.22.1
1.9
1.7
1.9
2.1
2.0
2.12.01.9
2.1
1.82.0
2.02.11.71.7
2.51.9
2.01.9
3.1
Magnetite(%)
2.04.3
3.0
3.8E-4
2.8
1.52.6
5.14.72.7
1.4
3.0
2.8
2.2
1.5
2.30.931.5
1.8
4.82.9
1.92.61.10.68
2.01.7
2.92.0
3.3
Tc ave.(°C)
560
579
579
57S
580
581
575
Primary direction (°)
Inc.
31715164.1
163.8
9.726
1
171151
-108304-47-54.2
68
161.5-143
130-74292-45.3
70106173
-750
126.1211-27
-154.5
59-115
232203165.2
1.54
959422.7
10620
331-153.8
9793
-25.445
233214-5.7
17015260
-140.8
Dec.
31
35
36
9.75859
4637
45343022.540
32.645
273932.1431118
15.6114.5
135835.5
3817106
13.730.228.1
33032.33435
-10-13.2
413122.7
-20-17
-7-27.2
303418
-16
Lithology
DuniteHarzburgiteHarzburgiteHarzburgiteDuniteTroctoliteOlivine gabbroHarzburgiteDuniteDuniteHarzburgiteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteGabbro (rodingite)DuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteDuniteHarzburgiteHarzburgiteDuniteDuniteDuniteDuniteHarzburgiteHarzburgiteDuniteDuniteDuniteDuniteDuniteHarzburgiteHarzburgiteHarzburgiteHarzburgiteDunite
Measuringlab
TAMUJRUWA, IRMIRMUWA, IRMJRTAMUUWA, IRMTAMUJRIRMTAMUJRIRMJRIRMJRUWA, IRMTAMUTAMUJRUWA, IRMJRIRMJRTAMUUWA, IRMTAMUTAMUTAMUIRMUWA, IRMJRUWA, IRMTAMUJRUWA, IRMUWA, IRMTAMUJRIRMIRMUWA, IRMUWA, IRMUWA, IRMIRMJRUWA, IRMTAMUJRIRMTAMUUWA, IRMJRIRMUWA, IRMIRMTAMUTAMUUWA, IRMJRIRMTAMUUWA, IRMUWA, IRM
-αX
rr
oo
op
nXrr
L,rπ"0>
2ööc
ROCK MAGNETIC PROPERTIES, SITE 895
it
I!
"J ^ü '—) O iJ O
'c 'c 'S '3 '3 '33 3 3 3 3 3
QQQQQQ
V) O ~H
"T T "T
00 00 00 00 00 00
inooiηoo-πq oq -* pN d m <N N' d r-icjrj
f^j^D^too^t^D ON 00 ON
c - intncnm CN ~- --<
O c o o c N ^ o o o ON o ~-<
^ H ^ H ^ H ^ ^ ^ H ^ H O O -*
0 0 c*~)^too^t^t-csim m m c*iOION x vo r•i in o o oon- vo *o -3- >o r- Nmmpooooo popö ö ö ö ö ö ööö
áááááá000000000000
3 c
δ Eε aii 2
|
S u
so p ooc < .38 « 5
60 II
ε ë
•^ • S H
2 °
2 - ε
js II 15
a = c.H bo£u g IISO <-> >
ε g sH ε §
h i> fi
Ml^p
illC >< !βo u x•a ^ u<β c H
| 1 IIC a p
IISIII
. . O cβ
7,
0.8
ç 0.6
I 0.4co
0.2
-Section 147-895D-2R-1.25 cm
Section 147-895D-3R-1.30 cm
Section 147-895D-6R-1.99 cm
— -Section 147-895E-6R-5.95 cm
Section 147-895E-7R-1.135 cm
50 100 150 200Temperature (K)
250 300
Figure 3. Saturation isothermal remanent magnetization given at 15 K(SIRM[15K]) vs. temperature in a nominally zero field (less than a few mil-litesla) with the Verwey transition of magnetite evident at l 18 K.
-Section 147-895D-2R-2.48 cm V-Section 147-895D-4R-4,75 cm *
Section 147•895D-5R-1,120 cmSection 147-895E-6R-2.67 cm
100 200 300 400Temperature (°C)
500 600
Figure 4. Thermal demagnetization of representative NRMs, normalized bytheir initial intensity. After the 400 °C measurement, the samples were storedin zero field and remeasured, and then thermal demagnetization was contin-ued.
als, dominantly magnetite, are secondary and are formed during theserpentinization process.
Magnetization Intensity
Assuming that magnetite is the only mineral that contributes toroom-temperature hysteresis, the amount of magnetite can be calcu-lated from the saturation magnetization extrapolated to zero field.The percentage of magnetite in individual samples is the of / s of thesample divided by Js of pure magnetite (92 Am2/kg) multiplied by100 (Table 1). The median magnetite concentration in the dunites andharzburgites is 2.0% and 1.0%, respectively.
Intensity values for the initial NRM are recorded in Table 1. Themedian NRM intensity of the dunites and harzburgites is 3.2 and 1.2A/m, respectively. A histogram of the NRM distribution is shown inFigure 6. We argue below that the samples have no significant drill-ing overprint; thus, the measured NRM should approximate the insitu magnetization. The observed magnetizations are similar in mag-nitude to those measured for oceanic basalts of seismic Layer 2a (e.g.,Johnson and Pariso, 1993) and serpentinized peridotites from other
409
P.R. KELSO, C. RICHTER, J.E. PARISO
0.5
0.4
0.3
0.2
0.1
n
SingleDomain
-
-
Pseudo-singleDomain
i
• Dunites
A Harzburgites
• Gabbronorites
-
-
-
Multidomain
• •
3Hc r / Hc
B
0.25
0.2
Jrs 1 Js
0.15
0.1
0.05
n
HA
A A
AA A •_
# i t A A• *
• Dunites
A Harzburgites "
-
•-
A
1.5 2 2.5• 'er ' 'c
Figure 5. A. Day et al. (1977) diagram of the hysteresis properties of Leg 147samples showing that the gabbronorites generally have a coarser bulk mag-netic grain size than the peridotites. B. Expansion of a portion of the Day etal. diagram showing that on average the harzburgites have larger JrJJs andlower HJHC ratios than the dunites.
ocean drilling studies (e.g., Dunlop and Prévot, 1982; Smith and Ban-erjee, 1985; Hamano et al., 1990).
The dunites and harzburgites have median low field magnetic sus-ceptibilities of 0.053 and 0.023 S1, respectively. Values for individualsamples are recorded in Table 1 and their distribution is shown in Fig-ure 7. These values are similar to those previously reported for ser-pentinized peridotites recovered from the ocean floor (e.g., Dunlopand Prévot, 1982; Hamano et al., 1990). Temperature- and field-de-pendent susceptibility measurements show that the low field suscep-tibility is nearly entirely due to magnetite (Richter et al., this volume).
There is a general inverse relationship between the degree of ser-pentinization and density within peridotites because serpentine is lessdense than its parent mineral, typically olivine and pyroxene (e.g.,Toft et al., 1990). A plot of magnetic susceptibility vs. density for theharzburgites shows that susceptibility increases as density decreases(Fig. 8). This relationship suggests that the degree of serpentinizationof the harzburgites may be an important factor contributing to themagnetic properties of oceanic peridotites. The highly serpentinizeddunites fall within a relatively restricted density range. Thus, theyshow only a weak inverse correlation of susceptibility with density.Similar correlations were observed between the NRMs and densitiesfor both the harzburgites and dunites. Highly serpentinized harzburg-ites have susceptibilities and densities comparable to those of the
30
25
1 0 -
5 -.
"i'0
ID Dunites (median = 3.2) ;
[3 Harzburgite (median = 1.2)"
y.•yy.•t••.•y>•K•i IW•&<ti
NRM (A/m)10 12 14
Figure 6. Histogram of the natural remanent magnetization of the dunites andharzburgites from Holes 895D and 895E.
Duntites (median = 0.053)
Harzburgites (median = 0.023)
0.06 0.08Susceptibility (SI)
0.12 0.14
Figure 7. Histogram of the magnetic susceptibility of the dunites andharzburgites from Holes 895D and 895E.
dunites, suggesting that differences in magnetic properties betweenthe harzburgites and dunites may, in large part, be due to differencesin their average degree of serpentinization.
The median Koenigsberger ratio (Q) for both the dunites andharzburgites is 2.4, which means that the median NRM of the sam-ples is 2.4 times larger than their induced magnetization (/,). The in-duced magnetization was calculated by multiplying the magneticsusceptibility times the present geomagnetic field at Site 895 (3.3 ×I0"5 T). The values of J, and Q for individual samples are recorded inTable 1. Site 895 Q values are similar to those found for serpentinizedoceanic peridotites by Dunlop and Prévot (1982) but nearly twice ashigh as values of serpentinized peridotites from ODP Hole 670A (Ha-mano et al., 1990).
The effect of temperature on the magnetic properties of these sam-ples is an important factor since rocks residing at depth in the oceaniclithosphere are at elevated temperature. The NRM decreases withtemperature (Fig. 4), whereas the magnetic susceptibility of PSD-MD magnetite remains relatively constant with temperature up tonear the Curie temperature (Richter et al., this volume). Therefore,the relative contribution of induced magnetization to magnetic anom-alies increases with temperature. These peridotites contain a signifi-cant amount of serpentine, a hydrous mineral that is not stable atelevated temperatures (>500°-550°C). Therefore, as temperaturesincrease, serpentine begins to break down, causing changes in themagnetic mineralogy and/or the concentration of magnetic minerals.
410
ROCK MAGNETIC PROPERTIES, SITE 895
0.1
α.Φo
0.01
— •- - Dunites
Harzburgites
147-895D-4R-1,41 cm
N, up
147-895D-6R-1,99cm
N, up
2.3 2.4 2.5 2.6 2.7 2.8 2.9Density (x1000 kg/m3)
Figure 8. Magnetic susceptibility is inversely related to density for the vari-ably serpentinized harzburgites. Linear regressions through the data for thedunites (dashed line) and harzburgites (solid line) have correlation coeffi-cients of 0.005 and 0.416, respectively. The dunites are all highly serpenti-nized and therefore show only a weak correlation. At lower densities (i.e.,higher degrees of serpentinization) the susceptibilities are similar for both theharzburgites and dunites.
Paleomagnetism
Petrologic and geochemical studies suggest that the serpentiniza-tion at Site 895 occurred at 350-450°C (Mével and Gillis; Früh-Green et al.; Mével and Stamoudi, all this volume), which is withinthe blocking temperature spectrum of magnetite. Therefore, the mag-netization of the samples is part thermal remanent magnetization(TRM) and part chemical remanent magnetization (CRM). Thisgrowth in CRM will be along the ambient magnetic field at the timethe grains grow through the critical volume, that is, above the super-paramagnetic/single-domain threshold (e.g., Haigh, 1958). Thus,whether the magnetization is TRM or CRM, it will record the direc-tion of the ambient geomagnetic field direction at the time the mag-netization was acquired. There is currently no paleomagnetictechnique to distinguish CRM from TRM if they have similar coer-civity spectrums.
Examples of orthogonal vector diagrams (Zijderveld, 1967) forsamples that have been either AF or thermally demagnetized areshown in Figure 9. Many of the samples display more than one mag-netic component, but the primary component is easily recognizable.The least-squared fit of this component is recorded in Table 1 as the"primary component." Both AF and thermal techniques were equallysuccessful in separating the individual components of magnetization.Therefore, most samples were AF demagnetized to eliminate anypossibility of chemical alteration and to preserve them for futurestudies. The softest component often has a steeper inclination thanthe primary component. Because Site 895 rocks have always been lo-cated near the magnetic equator, it is unlikely that the soft componentwas produced by the geomagnetic field. It has been suggested (e.g.,Johnson, 1978; Pariso and Johnson, 1993) that the drill string ofteninduces a soft, near-vertical component of magnetization. We sug-gest that this generally soft, weak, high-inclination component ofmagnetization is drilling-induced.
The samples are fairly stable magnetically, with the hardest com-ponent often being the dominant one. The median destructive field(MDF) is a way to estimate the ease of AF demagnetization and is theAF field required to demagnetize half of the NRM. The MDF is 10.9mT for the dunites and 13.1 mT for the harzburgites (Fig. 10). Theperidotites are much softer magnetically than the gabbronorites ofHole 894G, which have an average MDF of 31 mT (Pariso et al., this
Tick interval: 200 mA/m
147-895D-4R-4, 75 cm
N, up
Tick interval: 1000 mA/m
147-895D-6R-2,67cm
N, up
Tick interval: 200 mA/m Tick interval: 2000 mA/m
Figure 9. Zijderveld plots from AF (upper diagrams) and thermal demagneti-zation (lower diagrams). Open circles denote the horizontal component (E,N) and solid circles the east-west vs. the vertical (up-down) component. AFdemagnetization steps were NRM, 1, 2, 4, 6, 8, 10, 13, 16, 20, 25, 30, 35, 40,50, 60, 80, and 100 mT. Thermal demagnetization steps were NRM, 100°,200°, 300°, 400°, 450°, 500°, 525°, 550°, 575°, and 600°C.
volume), but are harder than the serpentinized peridotites of otherstudies that measured MDFs of <IO mT (e.g., Dunlop and Prévot,1982; Hamano et al., 1990). The samples have a wide range of un-blocking temperatures (Fig. 4), which is typical of relatively coarse-grained magnetite.
Because independent information on the horizontal orientation ofthe core does not exist, determination of the natural declination of themagnetization is not possible. Assuming that specimens originallyhad the same declination, it is possible to reorient individual seg-ments of cores relative to one another using the magnetic declination.This was done successfully during the analysis of the fabric andstructure orientations (MacLeod et al., Richter et al., both this vol-ume).
Viscous Remanent Magnetization
One potential problem that often complicates the interpretation ofpaleomagnetic data is the acquisition of a viscous remanent magneti-zation (VRM). The tendency of Site 895 samples to acquire a short-term (1 s to a few days) VRM was determined through room-temper-ature acquisition and decay experiments. Eight samples from Site895 with an unaltered NRM were stored in a field-free space for 54
411
P.R. KELSO, C. RICHTER, J.E. PARISO
20
15
5- 10
M Dunites (median = 10.9)E3 Harzburgites (median = 13.1)
-80Inclination (c
0
10 15MDF (mT)
20 25
Figure 10. Histogram of the median destructive fields for the dunites andharzburgites from Holes 895D and 895E.
days. The magnetization was measured periodically during this time.The change in the direction of magnetization was never more than afew degrees, which may largely be due to slight misorientations ofthe samples in the magnetometer. The intensity varied by less than1%. The one exception to this was Sample 147-895E-1R-2, 129-131cm, a highly magnetic dunite. This sample had a nearly 10% decreasein the NRM intensity, but there was negligible change in its direction.In the acquisition experiments, five samples with an unaltered NRMwere placed for 131 days in a 0.1-mT field (3 times the present-daygeomagnetic field at Site 895). The magnetization of the samples wasmeasured periodically by removing them from the field, making themeasurement in zero field, and then replacing them, oriented as be-fore, in the applied magnetic field. The field was oriented along theX axis (north) of the samples. NRM intensities changed by approxi-mately 10% or less during the more than 4 months of this experiment.We conclude, from these experiments and from the steep magneticinclinations observed in many samples, that if a room temperatureVRM exists it does not dominate the magnetization, although it maybe responsible for a secondary component of magnetization. Theseresults are consistent with the viscous induced magnetization (VIM)study by Pozzi and Dubuisson (1992), who found that the VIM ofHess Deep serpentinites is negligible at room temperature but is moreprevalent at elevated temperatures.
DISCUSSION
We have demonstrated that the Site 895 samples have a stablemagnetization and that the inclination of the remanence should be theinclination of the geomagnetic field at the time the magnetization wasacquired. Variation in the magnetic inclination within Holes 895Dand 895E was examined by plotting the inclination vs. depth (Figs.11, 12). These samples are relatively young (<l Ma) and have alwaysresided near the equator in a shallowly inclined magnetic field(<20°), even taking into account normal secular variations. Only dur-ing a magnetic reversal might the field be temporarily in a more in-clined orientation. Serpentinization is a metamorphic process thatoften occurs over an extended time period. As metamorphism pro-ceeds, magnetite grains gradually grow through their blocking vol-umes, with individual grains locking in their magnetization atdifferent times. Within a single sample, magnetization is acquiredover a long enough period of time to produce a time-averaged direc-tion for the Earth's magnetic field. Thus, polarity transitions and sec-ular variations in the field direction would be averaged out.
100
Figure 11. Stable paleomagnetic inclinations plotted as a function of depth(meters below seafloor) for Hole 895E. The relatively consistent inclinationsin the upper 55 m of the hole are more steeply dipping than the in situ geo-magnetic field at this site. This is likely caused by tectonic rotation of acoherent block after magnetization. The decrease in inclination toward thebottom of the hole may be due to tectonic rotations or a deviation of the drillhole from vertical.
0 r -
20
I 40E,.c
§"60
80
-
-
-
-
-40
•
m
Inclination0
•
m••
n
•
•
40
•
•t
i i
80• i
-
-
Figure 12. Stable paleomagnetic inclinations plotted as a function of depth(meters below seafloor) for Hole 895D. The lack of consistent inclinationdownhole suggests that the local region has experienced differential rotationsof blocks of a few meters or less in size. These rotations may be the result ofa very complicated tectonic history or, more likely, the hole was drilled in alandslide deposit, which caused the differential rotation of small blocks.
Figures 11 and 12 and Table 1 show that the stable magnetic in-clination is often >20°, which can be explained by sample rotation af-ter acquisition of the magnetization. Figure 11 shows a generaldecrease in inclination with depth in Hole 895E. The average inclina-tion in the upper portion of the hole is about +30° and it remains rel-atively constant throughout the upper 55 m. Below this depth theaverage inclination begins to shallow and becomes negative at thebottom of the hole. This trend is possibly caused by tectonic rotations(e.g., on listric faults), by differential rotation of individual blocks, or
412
ROCK MAGNETIC PROPERTIES, SITE 895
by changes in the drill-hole dip with depth. The latter explanation re-quires a change in the dip within Hole 895E by approximately 40°,which is unlikely. No logging was performed at Site 895 so there isno independent information on the dip of the drill hole. The magneticinclination in the upper portion of the hole is significantly greaterthan expected for an equatorial latitude. It is likely that Hole 895E un-derwent tectonic displacement that included at least 20° of coherentrotation around a horizontal axis.
The magnetic inclinations of the samples from Hole 895D rangefrom +60° to -60° with no downhole trends evident (Fig. 12). As ar-gued above, the samples had an initially shallow inclination; thus, thewidely scattered magnetic inclinations require differential rotation ofrelatively small blocks. It may be possible to explain the observed in-clination variations as a complex sequence of tectonic rotations; how-ever, an easier explanation is that this hole was drilled into alandslide, which caused the differential rotation of small individualblocks. It is possible that although the blocks have experienced dif-ferential rotations their relative positions may not have changedmuch. Thus, the initial stratigraphy within the hole may or may nothave been seriously disrupted.
CONCLUSIONS
Nearly pure pseudo-single-domain magnetite is the dominantmagnetic mineral in the peridotites of Site 895. The serpentinizeddunites and harzburgites have median NRMs of 3.2 and 1.2 A/m, andmedian induced magnetization of 1.4 and 0.61 (S1 units), respective-ly, which are similar to those of oceanic basalts, dikes, and gabbros.The dunites are more magnetic than the harzburgites, which may inpart be a result of their higher degree of serpentinization. Assumingthat serpentinized peridotites occur in the oceanic lithosphere, theyare a potential contributor to marine magnetic anomalies observed atthe ocean surface. The NRMs of the Site 895 samples are relativelystable and presumably a more important contributor to observedmagnetic anomalies than variations in the induced magnetization ofsimilar in situ rocks. However, serpentinized peridotites will have amore diffuse reversal boundary than the quickly cooled basalts ofLayer 2a due to the length of time over which the magnetization wasacquired. Serpentinized peridotites are even more likely a contributorto the longer-wavelength magnetic anomalies. The contribution ofserpentinized peridotites to magnetic anomalies depends critically onthe timing, length of time, degree, and depth extent of the serpentini-zation process and its resulting magnetization. A full understandingof these processes will require more studies of rocks from the upper-most mantle collected from representative environments in the oceanlithosphere, along with experimental studies on the process of mag-netization during serpentinization.
ACKNOWLEDGMENTS
We wish to thank the Institute for Rock Magnetism (IRM) at theUniversity of Minnesota where the majority of this work was con-ducted. The IRM is funded by the Keck Foundation, the National Sci-ence Foundation, and the University of Minnesota. We thankreviewers Paul Toft and Joseph Meert for helpful comments on theinitial draft of this manuscript. This work was supported by USSSPgrants to P.K. (147-20719b) and J.P. and a Deutsche Forschungsge-meinschaft grant to C.R. (R1576/1-2).
REFERENCES
Bina, M.M., and Henry, B., 1990. Magnetic properties, opaque mineralogyand magnetic anisotropies of serpentinized peridotites from ODP Hole670A near the Mid-Atlantic Ridge. Phys. Earth Planet. Inter., 65:88-103.
Bina, M.M., Henry, B., and Cannat, M., 1990. Magnetic anisotropy andsome other magnetic properties of serpentinized peridotites from ODPHole 670A. In Derrick, R., Honnorez, J., Bryan, W.B., Juteau, T., et al.,Proc. ODP, Sci. Results, 106/109: College Station, TX (Ocean DrillingProgram), 263-267.
Day, R., Fuller, M., and Schmidt, V.A., 1977. Hysteresis properties of titano-magnetites: grain-size and compositional dependence. Phys. EarthPlanet. Inter., 13:260-267.
Dunlop, D.J., and Prévot, M., 1982. Magnetic properties and opaque miner-alogy of drilled submarine intrusive rocks. Geophys. J. R. Astron. Soc,69:763-802.
Francheteau, J., Armijo, R., Cheminée, J.L., Hekinian, R., Lonsdale, P.F.,and Blum, N., 1990. 1 Ma East Pacific Rise oceanic crust and uppermostmantle exposed by rifting in Hess Deep (equatorial Pacific Ocean). EarthPlanet. Sci. Lett, 101:281-295.
Frost, B.R., 1985. On the stability of sulfides, oxides, and native metals inserpentine. /. Petrol, 26:31-63.
, 1991. Magnetic petrology: factors that control the occurrence ofmagnetite in crustal rocks. In Lindsley, D.H. (Ed.), Oxide Minerals: Pet-rologic and Magnetic Significance. Mineral. Soc. Am., Rev. Mineral.,25:489-509.
Gillis, K., Mével, C, Allan, J., et al., 1993. Proc. ODP, Init. Repts., 147:College Station, TX (Ocean Drilling Program).
Haigh, G., 1958. The process of magnetization by chemical change. Phil.Mag., 3:267-286.
Hamano, Y., Bina, M.M., and Krammer, K., 1990. Paleomagnetism of theserpentinized peridotite from ODP Hole 670A. In Detrick, R., Honnorez,J., Bryan, W.B., Juteau, T., et al., Proc. ODP, Sci. Results, 106/109: Col-lege Station, TX (Ocean Drilling Program), 257-262.
Johnson, H.P., 1978. Paleomagnetism of igneous rock samples—DSDP Leg45. In Melson, W.G., Rabinowitz, P.D., et al., Init. Repts. DSDP, 45:Washington (U.S. Govt. Printing Office), 387-396.
Johnson, H.P., and Pariso, J.E., 1993. Variations in oceanic crustal magneti-zation: systematic changes in the last 160 million years. J. Geophys. Res.,98:435-445.
Krammer, K., 1990. Rock magnetic properties and opaque mineralogy ofselected samples from Hole 670A. In Detrick, R., Honnorez, J., Bryan,W.B., Juteau, T., et al., Proc. ODP, Sci. Results, 106/109: College Sta-tion, TX (Ocean Drilling Program), 269-273.
Lonsdale, P., 1988. Structural pattern of the Galapagos microplate and evolu-tion of the Galapagos triple junction. J. Geophys. Res., 93:13551-13574.
Pariso, J.E., and Johnson, H.P., 1993. Do lower crustal rocks record reversalsof the Earth's magnetic field? Magnetic petrology of gabbros from OceanDrilling Program Hole 735B. J. Geophys. Res., 98:16013-16032.
Pozzi, J.P., and Dubuisson, G., 1992. High temperature viscous magnetiza-tion of oceanic deep crustal- and mantle-rocks as a partial source forMagsat magnetic anomalies. Geophys. Res. Lett., 19:21-24.
Prévot, M., Lecaille, A., and Mankinen, E.A., 1981. Magnetic effects ofmaghemitization of oceanic crust. J. Geophys. Res., 86:4009-4020.
Smith, G.M., and Banerjee, S.K., 1985. Magnetic properties of plutonicrocks from the central North Atlantic Ocean. In Bougault, H., Cande,S.C., et al., Init. Repts. DSDP, 82: Washington (U.S. Govt. PrintingOffice), 377-383.
Toft, P.B., and Arkani-Hamed, J., 1992. Magnetization of the Pacific Oceanlithosphere deduced from MAGSAT data. J. Geophys. Res., 97:4387-4406.
, 1993. Induced magnetization of the oceanic lithosphere andocean-continent magnetization contrast inferred from Magsat anomalies.J. Geophys. Res., 98:6267-6282.
Toft, P.B., Arkani-Hamed, J., and Haggerty, S.E., 1990. The effects of ser-pentinization on density and magnetic susceptibility: a petrophysicalmodel. Phys. Earth Planet. Inter., 65:137-157.
Zijderveld, J.D.A., 1967. AC demagnetization of rocks: analysis of results.In Collinson, D.W., Creer, K.M., and Runcorn, S.K. (Eds.), Methods inPaleomagnetism: New York (Elsevier), 254-286.
Beske-Diehl, SJ., 1990. Magnetization during low-temperature oxidation ofseafloor basalts: no large scale chemical remagnetization. J. Geophys.Res., 95:21413-21432.
Date of initial receipt: 2 August 1994Date of acceptance: 11 January 1995Ms 147SR-026
413