of basements rocks from dsdp leg 37' r.d. hyndman

11. SEISMIC VELOCITY MEASUREMENTSOF BASEMENTS ROCKS FROM DSDP LEG 37'

R.D. Hyndman, Victoria Geophysical Observatory, Earth Physics Branch,Department Energy, Mines, and Resources, Victoria, B.C., Canada

ABSTRACT

The seismic compressional and shear wave velocities have beenmeasured up to 2 kbar confining pressure for a large collection ofbasalts and a few gabbros and serpentinized peridotites recoveredon DSDP Leg 37. The mean compressiorial velocity for 79 basalts is5.94 km/sec. These results and the drilling record explain the lowvelocity upper crustal layer 2a sometimes outlined by seismic refrac-tion measurements. In these areas, there is an upper layer of highlyfractured low density volcanic material, flow breccia, drainedpillows, and perhaps lava tubes with extensive voids and some inter-calated sediments. The results also explain why the refractionvelocities for layer 2 are in general so much lower (5.0 km/sec) andmore variable than the velocities of little-weathered sea-floor basaltsmeasured in the laboratory (6.0 km/sec). The refraction velocitydepends on the amount of fracturing, voids, and sediment and noton a downward change in general rock composition ormetamorphic grade.

Poisson's ratio for little-weathered basalts is found to decreasesystematically with decreasing velocity. The relation predicts aPoisson's ratio of 0.28 for a layer 2 refraction compressional veloci-ty of 5.0 km/sec and 0.24 for a layer 2a velocity of 2.8 km/sec ingood agreement with the values observed. Poisson's ratio fromrefraction measurements should distinguish between low velocityfractured basalt and similar velocity sediments since the lattershould have high Poisson's ratios.

One hole penetrated a complex section of gabbros, serpentinizedperidotites, and breccias. The gabbros have a mean velocity of 7.21km/sec, appropriate to the upper part of crustal layer 3 (i.e., 3a) if asmall amount of fracturing or low velocity material is present. If theultramafics in the section were little serpentinized before beingbrought to near the surface, such a complex at the base of the crustwould explain the basal layer (3b) that has sometimes been observedby refraction measurements with velocities between those of oceaniclayer 3 and the mantle.

INTRODUCTION

Seismic refraction measurements provide the mostimportant presently available information on the struc-ture of the oceanic crust. The interpretation of thesemeasurements in terms of the rock compositions thatare present in the crust requires a comparison of therefraction velocities with velocities measured in thelaboratory for a wide range of rock types. Much impor-tant information has been obtained by measurementson dredged sea-floor rocks, but far more representativeand less weathered samples from greater depths in thecrust are now being obtained by the Deep Sea DrillingProject. Leg 37 provided samples from over 500 metersdepth into layer 2, five times the previous maximum.

'Contribution 597 of the Earth Physics Branch.

This depth is sufficient for the first time, to compare thevelocity-depth profile in layer 2 by refractionmeasurements over the drill sites with velocitiesmeasured on rocks from the drill holes. Unfortunatelyno in-hole velocity measurements could be made, butlaboratory measurements on a large number of samplesare reported here.

Laboratory velocities have been measured on a largenumber of samples previously obtained by dredgingand drilling. A summary of the data is provided in thereview by Christensen and Salisbury (1975). Details ofmany of the measurements have been reported inprevious volumes of the Initial Reports of the Deep SeaDrilling Project.

Direct sampling by dredging and drilling and correla-tion with ophiolite complexes have shown that uppercrustal layer 2 must consist primarily of extrusivebasalt. The mean and standard deviation of seismic

373

R. D. HYNDMAN

refraction velocities for layer 2 is 5.04 ±0.69 km/sec(Christensen and Salisbury, 1975; also Raitt, 1963;Ludwig, et al., 1971). Laboratory measurements ofbasalts give a mean velocity of about 5.5 km/sec(Christensen and Salisbury, 1975) which is much higherthan the mean, but within the range of refractionvelocities. However, many of the measured sampleshave undergone extensive sea-floor weathering that hasconsiderably reduced their velocity. The weathering ex-tends only to a shallow depth so these samples are notrepresentative of the major thickness of layer 2.Christensen and Salisbury (1972, 1973) show that theweathering and reduction in velocity increaseprogressively with age, and Hyndman (1974) andSalisbury and Christensen (1976) showed that theweathering reduction in velocity decreases rapidly withdepth. Even in 100-m.y. old sea floor, the reducedvelocity extends to a depth of only 50 meters. Thus,only in quite old sea floor is the weathered layer of suf-ficient thickness to be important in seismic refractionmeasurements (although the increasing thickness of theweathered layer may explain the small decrease inrefraction velocity with age that has been suggested forlayer 2, e.g., Christensen and Salisbury, 1972; Hart,1973). Young sea floor basalts (less than 20 m.y.) andthose from deep holes into older sea floor have meanvelocity of about 6.0 km/sec (e.g., Christensen andSalisbury, 1975). Thus, there is a major discrepancybetween refraction and laboratory velocities that re-quires large-scale (larger than laboratory samples) frac-turing, intercalated sediment, volcanic breccias, or rub-ble in layer 2. The discrepancy is even larger in someregions such as the Reykjanes Ridge (Talwani et al.,1971) where the upper part of layer 2 (designated layer2a) has velocities between 2.3 and 3.7 km/sec.

The composition of lower crustal layer 3 is still muchdebated since probable samples have been obtainedonly where they have been exposed by tectonic motionsuch as along fractures. The most probable rocks aregabbro, metagabbro, metabasalt, and possiblyamphibolite. All give reasonable agreement with themean seismic refraction velocity of 6.73 ±0.19 km/sec.Although fresh gabbros have somewhat higherlaboratory velocities, the layer velocity may be reducedby fracturing and low velocity material in the sameway, although much less, as is apparent for layer 2. Thelowering of velocities by fracturing is illustrated by thevery low seismic refraction velocities found by Lort andMatthews (1972) in the gabbro section of the Troodosophiolite complex. Shear velocities from refractionmeasurements for layer 3 are few but suggest Poisson'sratios from 0.24 to 0.31. The above rocks havelaboratory values in this range; gabbro about 0.30 andmetabasalt about 0.28. Serpentinites have much higherPoisson's ratios of about 0.37 so they cannot be a majorconstituent of layer 3 (Christensen, 1972; Christensenand Salisbury, 1975).

VELOCITY MEASUREMENTSThe velocity measurements reported here were made

with two different instruments onboard GlomarChallenger. Most samples were measured, first at one

atmosphere pressure on the Hamilton frame apparatus,then to 2.0 kbar in a high pressure apparatus. The es-timated accuracies of the techniques are: Hamiltonframe, compressional waves only, ±3%; shipboardmeasurements to 2.0 kbar, compression ±1%, shear±2%. Data are given in Table 1.

The Hamilton frame apparatus (Hamilton, 1965;Cernock, 1970) provides a velocity at atmosphericpressure that cannot easily be obtained on- the highpressure equipment as well as providing a check onmajor errors in the high pressure velocities. The ap-paratus is designed for the rapid measurement ofsediments that have low velocity and thus long traveltimes through the sample. Very good time resolution isnot required. The transmitted pulse has a slow rise timeand the equipment uses an oscilloscope variable timedelay for time measurement which has fairly low ac-curacy. Also it is difficult to make good acoustic con-tact with igneous rock. The values given are the meansof values with the pulse travel in opposite directionthrough the length of the minicores.

Both compressional and shear wave velocities weremeasured onboard ship from 0.07 kbar (2500 psi) to2.07 kbar (30,000 psi) using a method essentially asdescribed by Birch (1960) and Christensen and Shaw(1970). The pressure was applied with a hydraulic fluidmedium which is excluded from the sample, using aHigh Pressure Equipment Ltd, Model LM-I. IsostaticPress pressure chamber with 10-cm inside diameter andpumping unit. The pressure can be measured to ±2%and temperature to ±0.2°C with an internal thermistor.The samples are 2.5 cm diameter and 3 to 5 cm long.The circumference of the sample is jacketed by a thinmetal sheet with a soldered joint to exclude the pressuremedium. Thus, the confining pressure is close to the ex-ternal pressure and the pore pressure is small; see com-ments below about relating laboratory confiningpressure to depth in the crust. The ends of the samplesare painted with silver-filled epoxy resin to make elec-trical contact with the pulse transducers and with thejacket. The transducers are 2.54 cm in diameter andabout 0.1 cm thick with a resonant frequency of 2.0mHz from Valpey Fisher Ltd. Barium titanatetransducers in compressional mode were used togenerate compressional waves and AC cut quartztransducers in transverse mode to generate shear waves.The transducers are fastened with electrically con-ducting epoxy to 2.60 cm diameter by 4.0 cm longstainless steel electrodes. Gum rubber tubing seals thespace between the electrodes and jacketed specimen.Brass cylinders around the whole assembly act as elec-tronic shields. All electrical connections are with shield-ed cables. A Hewlett-Packard Model 214-A pulsegenerator is used to generate 0.3 µsec wide pulses to ex-cite the transmitting transducer. Ringing is reduced byusing a 50-ohm termination to match the output im-pedance of the generator. High frequency cross-coupling between input and received pulse lines wasreduced by smoothing the output pulse with a 0.004microfarad capacitance. The pulse transit time throughthe sample is measured by comparison with a variablemercury-delay-line. The output pulse is split and sent

374

SEISMIC VELOCITY MEASUREMENTS OF BASEMENT ROCKS

TABLE 1Seismic Velocities

Sample(Interval in cm)

Hole 332A

6-2,117-120Basalt

7-1,66-69Basalt

7-2, 41-42Basalt

8-2, 6-9Basalt

12-1,122-125Basalt

16-1, 33-36Basalt

21-1, 131-134Basalt

27-1.128-131Basalt

28-1,83-86Basalt

28-3, 28-31Basalt

29-1,74-77Basalt

30-1,127-130Basalt

BulkDensity

(g/cm•3)

2.792

2.810

2.811

2.809

2.796

2.719

2.810

2,710

2.790

2.791

2.846

2.822

PorSor σ

P

S

σ

P

sσ

P

sσ

P

sσ

P

so

P

sσ

P

sσ

P

sσ

P

S

a

P

sσ

P

sσ

P

sö

Incr. orDeer.Press.

IDID

IDID

IDID

IDID

IDID

IDID

IDID

IDID

IDID

IDID

IDID

IDID

Ham.Frame

6.09

6.14

5.46

5.42

5.37

5.16

5.66

5.33

5.63

5.52

5.82

5.83

0.07

6.576.49

3.15

5.735.72

3.150.28

5.645.57

5.405.322.862.880.30

5.986.02

5.625.52

5.725.59

6.066.09

6.156.11

0.17

6.35

6.596.513.383.380.32

5.603.163.090.27

5.745.753.193.170.28

5.655.593.043.070.29

5.395.332.882.900.30

6.016.03

3.190.30

5.625.582.983.000.30

5.755.633.143.150.28

5.71

3.173.190.28

6.116.133.303.320.29

6.156.113.313.330.29

0.34

6.40

3.340.31

6.606.523.383.390.32

5.695.673.133.100.28

5.765.803.223.190.28

5.675.623.073.080.29

5.425.402.922.940.29

6.036.043.183.210.31

5.635.633.023.050.30

5.775.703.183.180.28

5.735.673.193.190.27

6.136.123.333.330.29

6.156.123.323.340.29

Velocity (km/sec)Pressure (kbar)

0.52

6.406.383.393.410.30

6.596.533.383.390.32

5.695.713.133.120.29

5.805.853.223.220.28

5.675.643.083.090.29

5.455.442.942.960.29

6.066.063.203.230.30

5.645.653.043.070.29

5.795.723.193.190.28

5.785.723.203.200.28

6.156.133.333.330.29

6.156.133.333.350.29

0.69

6.386.423.403.430.30

6.596.573.383.390.32

5.715.763.143.130.29

5.855.903.243.240.28

5.685.663.083.100.29

5.465.482.962.960.29

6.066.053.223.250.30

5.665.683.073.090.29

5.815.743.203.200.28

5.795.763.203.220.28

6.156.153.333.330.29

6.156.133.343.350.29

1.03

6.466.443.433.470.30

6.586.593.393.390.32

5.785.793.143.160.29

5.935.963.263.280.28

5.695.683.083.110.29

5.505.513.002.980.29

6.086.073.253.270.30

5.695.723.113.110.29

5.835.783.213.220.28

5.805.813.213.230.28

6.156.173.333.330.29

6.156.143.353.360.29

1.38

6.486.483.463.490.30

6.576.593.403.400.32

5.805.823.173.180.29

5.986.003.273.300.28

5.705.713.093.110.29

5.525.553.01

0.29

6.096.083.273.290.30

5.745.753.123.130.29

5.855.843.233.220.28

5.835.843.223.230.28

6.166.173.333.330.29

6.166.163.363.360.29

1.72

6.496.503.493.500.30

6.596.603.413.400.32

5.825.833.183.190.29

6.026.033.303.320.28

5.715.733.103.110.29

5.555.573.01

0.29

6.096.103.293.300.29

5.755.773.143.140.29

5.865.873.233.230.28

5.855.863.233.230.28

6.176.173.333.330.29

6.176.173.363.360.29

2.07

6.496.503.503.500.30

6.586.583.413.400.32

5.855.853.203.200.29

6.056.053.323.320.28

5.725.733.113.110.29

5.585.583.01

0.29

6.106.103.303.300.29

5.775.793.153.150.29

5.895.883.243.230.28

5.875.883.233.240.286.196.193.333.330.30

6.176.183.373.360.29

375

R. D. HYNDMAN

TABLE 1 - Continued


Hole 332A Cont.

31-2,32-35Basalt

31-3, 79-82Basalt

33-2, 92-95Basalt

33-2,128-131Basalt

34-1,88-91Basalt

34-2, 104-107Basalt

36-2, 33-36Basalt

37-1, 116-119Basalt

40-2, 95-98Basalt

40-3, 40-43Basalt

Hole 332B

1-5, 27-30Basalt

1-5,120-123Basalt

BulkDensity(g/cm3)

2.732

2.780

2.813

2.855

2.816

2.810

2.827

2.818

2.741

2.861

2.796

2.812

Por Sor σ

P

S

σ

P

sσ

P

sσ

P

sσ

P

sσ

P

sσ

P

S

σ

P

sσ

P

sσ

P

sσ

P

sσ

P

sσ

Incr. orDeer.Press.

IDID

IDID

IDID

IDID

IDID

IDID

IDID

I

DID

IDID

IDID

IDID

IDID

Ham.Frame

5.18

5.44

5.58

5.88

5.76

5.37

5.80

5.69

5.15

5.79

5.79

5.63

5.615.57

5.665.68

5.835.873.233.230.28

6.14

6.075.97

5.675.633.023.040.30

6.196.233.363.350.29

5.885.87

5.52

6.456.453.433.450.30

6.276.28

3.053.04

5.665.643.003.000.30

5.735.743.103.100.29

5.865.883.243.220.28

6.166.153.333.330.29

6.045.963.203.190.30

5.685.723.063.080.30

6.216.253.363.360.29

5.895.883.233.240.28

5.585.693.053.080.30

6.466.453.433.460.30

6.316.333.373.390.30

6.096.083.153.180.31

5.705.693.043.050.30

5.785.773.123.120.29

5.885.893.233.210.29

6.186.173.333.320.30

6.076.033.233.240.30

5.765.763.093.090.30

6.276.233.363.360.30

5.935.893.223.230.29

5.925.873.133.160.30

6.476.483.463.480.30

6.336.343.383.390.30

6.176.223.233.280.31


5.715.733.073.070.30

5.805.803.133.140.29

5.905.903.233.220.29

6.196.173.323.310.30

6.096.093.253.270.30

5.795.783.113.110.30

6.286.233.373.360.30

5.925.913.233.230.29

5.945.953.173.200.30

6.496.493.473.490.30

6.356.363.323.390.30

6.216.263.263.310.31

5.755.773.093.100.30

5.825.813.143.140.29

5.905.913.243.220.29

6.196.193.323.320.30

6.116.143.273.300.30

5.805.803.123.130.30

6.276.243.373.370.30

5.935.923.223.230.29

5.966.003.193.220.30

6.486.503.493.490.30

6.376.383.393.410.30

6.236.273.303.340.30

5.785.803.123.130.29

5.835.833.153.150.29

5.925.923.243.230.29

6.196.203.333.320.30

6.146.173.303.320.30

5.835.823.143.140.30

6.276.263.373.380.30

5.945.933.233.230.29

6.036.013.223.260.30

6.506.513.503.500.30

6.416.413.403.410.30

6.316.313.343.370.30

5.825.843.143.150.29

5.855.853.163.160.29

5.945.923.233.230.29

6.206.203.333.330.30

6.176.203.323.340.30

5.855.853.153.160.29

6.286.273.383.390.29

5.955.953.233.230.29

6.076.053.253.270.30

6.526.523.503.510.30

6.436.443.423.420.30

6.346.343.373.390.30

5.845.853.153.150.30

5.865.863.163.170.29

5.935.923.233.230.29

6.216.213.333.330.30

6.196.213.343.350.29

5.865.863.163.170.29

6.296.283.383.390.30

5.965.963.223.230.29

6.116.103.273.290.30

6.526.523.513.510.30

6.456.463.423.420.30

6.376.363.393.400.30

5.865.863.163.160.29

5.885.883.173.170.30

5.935.943.233.230.29

6.216.213.333.340.30

6.236.233.353.350.30

5.885.883.173.170.30

6.296.293.393.380.30

5.975.973.233.230.29

6.136.123.303.300.30

6.536.543.523.510.30

6.486.483.433.430.31

6.386.383.403.400.30

376


TABLE 1 - Continued


Hole 332B Cow?

2-1,60-63Basalt

2-2, 86-89

2-5, 69-72Basalt

2-5,115-117Basalt

3-2, 116-119Basalt

3-4,10-13Basalt

3-4, 7-20Basalt

6-1,100-103

6-2, 122-123Basalt

8-3, 22-24Basalt

9-1,112-115Basalt

9-2,104-106Basalt

9-3, 80-82Basalt

10-2,125-128Basalt

14-2, 81-83Basalt

15-1,129-131Basalt

BulkDensity(g/cnA

2.780

2.841

2.750

2.833

2.746

2.814

2.891

2.843

2.659

2.727

2.806

2.796

2.873

2.839

2.693

2.880

PorSor σ

P

S

σ

P

S

P

sσ

P

sσ

P

sσ

P

P

P

sσ

P

P

P

sσ

P

P

sσ

P

P

P

s

σ

Incr. orDeer.Press.

IDID

IDID

IDID

IDID

IDID

ID

ID

IDID

ID

ID

ID

ID

IDID

I

D

I

D

ID

I

D

Ham.Frame

5.94

5.76

5.91

6.03

5.91

6.09

6.13

5.89

6.16

5.40

5.25

4.99

5.89

5.08

4.83

5.69

0.07

6.396.433.29

0.32

6.166.25

6.46

3.55

0.28

3.473.51

6.086.083.293.270.29

6.42

5.945.923.18

0.30

5.58

5.43

2.912.950.29

5.405.43

6.316.35

5.695.68

5.41

5.985.793.243.250.28

0.17

6.506.543.313.320.33

6.246.283.393.410.29

6.526.433.583.610.28

6.54

3.503.520.30

6.126.093.313.310.29

6.486.39

6.436.41

6.035.963.193.220.30

5.385.43

5.745.60

5.485.542.962.990.29

5.485.48

6.426.40

5.675.64

5.46

5.965.923.253.260.29


0.34

6.556.573.323.330.33

6.326.323.393.410.30

6.546.513.613.620.28

6.566.613.523.530.30

6.136.113.323.310.29

6.496.48

6.466.47

5.995.963.203.240.29

5.455.49

5.825.87

5.565.583.003.030.29

5.555.57

6.416.423.223.240.33

5.745.67

5.49

5.985.983.273.270.29

0.52

6.586.593.333.340.33

6.356.353.413.420.30

6.556.553.633.620.28

6.596.613.523.520.30

6.126.143.333.320.29

6.526.51

6.466.48

6.005.983.213.250.30

5.525.51

5.835.90

5.625.633.033.050.29

5.595.60

6.436.443.233,270.33

5.775.70

5.52

6.016.013.293.290.29

0.69

6.606.593.343.350.33

6.366.373.423.430.30

6.566.583.643.620.28

6.616.593.533.520.30

6.146.153.333.330.29

6.546.55

6.496.50

6.006.003.233.260.29

5.555.55

5.905.92

5.635.693.053.060.29

5.625.65

6.436.453.263,320.32

5.815.75

5.56

6.046.043.293.300.29

1.03

6.616.613.363.360.33

6.396.393.423.430.30

6.606.603.643.630.28

6.656.593.533.510.30

6.166.173.333.340.29

6.576.56

6.516.53

6.026.033.253.280.29

5.605.61

5.985.91

5.695.733.073.080.30

5.695.70

6.436.453.323.330.32

5.815.82

5.595.56

6.066.073.303.310.29

1.38

6.636.633.363.360.33

6.426.413.443.440.30

6.626.613.653.640.28

6.646.613.533.520.30

6.176.183.343.350.29

6.586.58

6.536.55

6.046.053.263.280.29

5.645.64

5.955.95

5.745.763.083.090.30

5.735.74

6.446.463.333.330.32

5.845.86

5.625.58

6.086.093.313.320.29

1.72

6.656.653.373.370.33

6.446.433.443.440.30

6.656.633.663.640.28

6.656.633.533.520.30

6.196.203.353.350.29

6.606.59

6.546.56

6.076.073.273.290.29

5.686.67

5.975.97

5.775.773.103.100.30

5.765.78

6.456.473.353.330.32

5.865.88

5.635.61

6.116.103.323.320.29

2.07

6.676.673.373.370.33

6.456.463.453.440.30

6.666.653.663.650.28

6.646.643.533.520.30

6.206.213.353.350.29

6.606.61

6.576.58

6.086.093.293.290.29

5.705.70

5.996.00

5.795.793.113.110.30

5.795.80

6.476.473.353.340.32

5.905.91

5.635.63

6.126.123.333.330.29

377

R. D. HYNDMAN

TABLE 1 - Continued


BulkDensity(g/cm3)

Hole 332B Cont.19-1,104-107 2.871

Basalt

20-1,56-59Basalt

21-1,99-102Basalt

22-1,57-60Basalt

22-2, 140-142Basalt

22-3. 39-41Basalt

22-4,11-13Basalt

24-1,125-127Basalt

25-2, 91-93Basalt

25-4, 47-49Basalt

27-2, 93-95Basalt

28-2, 23-26Basalt

29-1, 86-89Basalt

31-1,108-111

33-1,62-65

35-1,66Basalt

36-1,77-79Basalt

36-3, 76-78Basalt

36-3, 143-145Basalt

36-4,131-133Basalt

36-5,40-42Basalt

36-6,44-46Basalt

37-3, 54-56Basalt

44-1,75-77Basalt

47-2, 145-147Basalt

2.832

2.900

2.850

2.614

2.862

2.877

2.731

2.876

2.829

2.841

2.829

2.879

2.736

2.835

2.813

2.788

2.780

2.809

2.696

2.783

2.866

2.883

2.609

2.88S

PorSor σ

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

P

Incr. orDeer.Press.

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

ID

Ham.Frame

5.41

5.31

5.57

5.68

4.40

5.62

5.87

5.27

5.43

5.48

5.75

5.40

5.28

5.00

5.37

5.61

5.61

5.14

5.37

4.98

5.44

5.64

5.35

5.07

5.91

0.07

5.625.64

5.38

5.725.73

5.895.84

4.684.81

5.765.84

6.136.19

5.53

5.495.53

5.825.72

5.89

5.515.39

5.75

5.31

5.73

5.54

5.305.42

5.72

5.705.71

4.725.04

5.645.64

0.17

5.615.71

5.45

5.775.74

5.915.88

4.784.88

5.845.89

6.186.21

5.605.60

5.525.56

5.805.75

5.995.95

5.545.46

5.835.89

5.365.33

5.895.87

5.886.15

5.825.87

5.81

5.805.82

5.455.51

5.745.62

6.126.05

5.685.73

5.045.11

5.685.66


0.34 0.52 0.69

5.745.81

5.53

5.805.79

5.955.93

4.895.00

5.945.99

6.216.23

5.665.68

5.575.61

5.845.84

6.016.06

5.565.57

5.905.94

5.435.41

5.945.93

6.246.48

5.885.90

5.925.90

5.885.86

5.515.52

5.755.70

6.086.16

5.785.82

5.135.19

5.715.72

5.815.83

5.585.51

5.825.83

5.975.93

5.025.09

5.976.02

6.236.23

5.765.75

5.585.62

5.865.89

6.036.07

5.615.63

5.955.96

5.465.44

5.935.97

6.446.57

5.925.94

5.935.95

5.905.92

5.535.53

5.785.75

6.156.20

5.835.85

5.195.24

5.745.79

5.845.86

5.615.54

5.845.85

5.985.96

5.095.12

6.036.05

6.256.25

5.765.80

5.625.67

5.905.93

6.076.08

5.635.65

5.966.00

5.475.48

5.946.01

6.526.60

5.955.96

5.945.95

5.955.93

5.545.54

5.785.78

6.186.23

5.865.90

5.235.26

5.815.84

1.03

5.905.90

5.675.56

5.875.90

5.985.97

5.135.14

6.096.09

6.276.27

5.845.84

5.665.72

5.955.95

6.116.10

5.655.66

6.026.00

5.495.50

5.996.02

6.596.62

5.996.00

5.985.98

5.985.97

5.555.56

5.825.85

6.236.25

5.935.94

5.295.31

5.845.86

1.38

5.925.93

5.725.67

5.905.92

6.005.98

5.175.16

6.126.13

6.286.29

5.865.87

5.705.76

5.996.00

6.136.10

5.665.68

6.046.04

5.515.52

6.026.03

6.646.65

6.016.02

6.016.00

6.005.98

5.585.56

5.855.88

6.286.29

5.985.97

5.305.34

5.865.88

1.72

5.955.97

5.745.75

5.925.95

6.006.01

5.21

5.21

6.156.15

6.306.30

5.915.92

5.765.80

6.036.03

6.116.11

5.695.70

6.066.07

5.535.53

6.056.05

6.676.65

6.036.04

6.026.01

6.036.01

5.595.58

5.875.90

6.336.32

6.036.01

5.335.36

5.895.90

2.07

5.975.99

5.765.79

5.955.96

6.036.02

5.235.23

6.166.16

6.316.30

5.945.93

5.815.82

6.066.06

6.106.10

5.715.71

6.096.09

5.545.54

6.076.07

6.706.70

6.056.05

6.036.02

6.046.03

5.605.59

5.915.92

6.346.34

6.046.03

5.355.37

5.925.92

378


TABLE 1 - Continued


Hole 333A

1-3, 74-76Basalt

3-1, 129-131Basalt

6-2, 58-61Basalt

8-6, 58-60Basalt

10-2,112-115Basalt

Site 334

16-2, 88-91Basalt

16-4, 104-107Basalt

18-1,84-87Basalt

18-2,12-14Basalt

21-1,39-41Gabbro

214, 78-92Gabbro

22-1,69-71Gabbro

22-2,43-45Serpentinite

23-1,76-78Gabbro

BulkDensity

(g/cm3)

2.849

2.821

2.872

2.769

2.763

2.820

2.928

2.945

2.893

3.002

2.969

3.013

2.836

3.034

PorSor σ

P

P

P

P

P

P

sσ

P

sσ

P

sσ

P

sσ

P

sσ

P

s

σ

P

sσ

P

sσ

P

sσ

Incr. orDeer.Press

ID

ID

ID

ID

ID

IDID

IDID

ID1D

IDID

IDID

IDID

IDID

IDID

IDID

Ham.Frame.

5.74

5.55

6.06

5.47

5.23

5.80

6.20

6.25

6.23

7.17

7.61

6.82

6.16

7.13

0.07

5.375.34

5.435.44

5.935.86

5.565.62

5.125.17

5.935.913.193.200.29

6.393.503.480.29

6.406.39

3.470.29

6.356.323.473.470.29

7.126.92

6.95

6.15

3.05

0.34

3.843.83

0.17

5.375.33

5.525.55

5.985.89

5.585.65

5.215.23

5.975.923.203.200.30

6.526.423.493.480.30

6.406.383.473.470.29

6.356.363.473.460.29

7.297.25

3.800.31

7.177.093.78

0.30

6.92

3.693.670.30

6.446.523.143.190.34

7.117.193.843.830.30


0.34

5.395.35

5.615.58

6.005.90

5.655.66

5.255.26

5.985.933.203.200.30

6.506.433.493.480.30

6.406.393.473.470.29

6.396.393.463.460.29

7.307.283.793.810.31

7.197.133.813.840.30

6.92

3.703.680.30

6.646.713.193.220.35

7.247.223.853.840.30

0.52

5.425.37

5.625.61

6.015.93

5.675.67

5.275.29

5.965.933.203.200.30

6.506.433.493.480.30

6.406.403.483.470.29

6.416.393.473.460.29

7.297.293.793.820.31

7.167.173.853.870.30

6.97

3.713.680.30

6.766.733.223.220.35

7.247.223.853.850.30

0.69

5.425.38

5.645.64

6.025.96

5.705.70

5.305.32

5.965.933.203.200.30

6.466.443.503.490.29

6.416.423.483.470.29

6.416.413.473.460.29

7.297.293.803.820.31

7.197.173.873.870.30

6.97

3.723.700.30

6.796.893.223.210.36

7.237.223.853.850.30

1.03

5.455.45

5.675.67

6.066.00

5.735.73

5.345.35

5.975.953.213.200.30

6.466.463.503.490.29

6.416.433.483.480.29

6.416.413.483.470.29

7.307.303.813.820.31

7.237.203.873.870.30

6.99

3.743.720.30

6.936.873.213.210.36

7.277.233.863.860.30

1.38

5.485.50

5.695.71

6.086.05

5.755.75

5.385.39

5.995.973.213.200.30

6.466.453.503.490.29

6.426.433.493.480.29

6.416.423.483.470.29

7.327.303.823.830.31

7.277.243.883.870.30

7.01

3.743.730.30

6.946.883.213.210.36

7.257.253.863.860.30

1.72

5.525.54

5.725.73

6.066.06

5.775.78

5.405.41

6.005.983.213.210.30

6.476.453.503.490.29

6.436.433.493.480.29

6.416.413.483.470.29

7.337.313.833.830.31

7.287.283.883.880.30

7.02

3.743.740.30

6.986.933.213.210.36

7.277.263.863.860.30

2.07

5.555.55

5.755.75

6.086.08

5.795.79

5.435.43

6.015.993.213.210.30

6.466.483.503.500.29

6.446.443.493.490.29

6.41

3.483.480.29

7.327.323.843.840.31

7.297.293.883.880.30

7.027.013.743.750.30

6.976.953.213.210.36

7.287.273.863.870.30

379

R. D. HYNDMAN

TABLE 1 - Continued


24-1,63-65Gabbro

24-4, 86-88Gabbro

Site 335

5-2, 36-38Basalt

6, ccBasalt

6-1,5-7Basalt

6-3, 44-49Basalt

6-4, 23-25Basalt

6-6, 75-77Basalt

BulkDensity(g/cm3)

2.871

2.851

2.866

2.753

2.822

2.800

2.519

2.789

PorSor σ

P

S

σ

P

s

a

P

P

P

P

P

P

Incr. orDeer.Press.

IDID

IDID

ID

ID

ID

ID

ID

ID

Ham.Frame

7.11

6.39

5.96

5.52

5.60

5.58

5.17

5.67

0.07

6.926.823.70

0.30

6.676.74

6.186.17

5.625.64

5.805.80

5.85

5.785.82

0.17

7.057.023.713.710.31

6.716.773.64

0.29

6.176.19

5.705.73

5.855.87

5.88

5.295.31

5.815.86


0.34 0.52 0.69 1.03

7.167.263.723.730.32

6.776.873.643.610.30

6.216.20

5.765.76

5.925.93

5.87

5.355.32

5.885.87

7.237.303.723.710.32

6.836.873.663.610.30

6.226.22

5.765.76

5.945.93

5.88

5.385.33

5.905.89

7.307.323.733.710.33

6.886.883.683.620.30

6.246.24

5.785.80

5.955.93

5.87

5.355.31

5.935.91

7.347.353.733.720.33

6.916.883.673.620.31

6.256.26

5.795.81

5.965.94

5.88

5.375.33

5.935.93

1.38

7.367.393.733.730.33

6.916.893.663.620.31

6.276.27

5.805.82

5.975.96

5.895.84

5.385.36

5.955.95

1.72

7.417.423.733.740.33

6.946.913.663.630.31

6.296.28

5.805.82

5.995.97

5.905.89

5.395.39

5.955.95

2.07

7.427.423.743.740.33

6.926.943.673.660.31

6.296.29

5.815.83

5.985.97

5.905.90

5.405.40

5.955.95

through the sample and the delay line and the tworeceived pulses displayed simultaneously on a dualtrace Telequipment D-83, 50-mHz oscilloscope. Themercury delay line consists of a column of mercury withone transducer fixed at the base and a second mountedon a screw thread so as to move vertically, changing thelength of mercury between the two transducers. Thelength, measured with a precision dial gage, is adjusteduntil the two received pulses are superimposed on theoscilloscope. The travel times through the sample andmercury then should be identical. As pointed out byBirch (1960) the mercury-delay-line not only has highprecision and stability of about 0.01 µsec, but also, theshape of the received pulse through the mercury issimilar to that through the sample. At least the firstquarter wavelength of the two pulses can be matched byadjusting the amplitude scales. Thus, the time uncer-tainty is much less than would result from trying tolocate the onset of the received pulse which, because ofelectrical and mechanical response times, builds upslowly and is followed by damped oscillations. Thedelay-line was calibrated with a high precision digitaltime interval unit. A calibration of 6.902 µsec/cm(17.53 µsec/in.) was used, which is identical to thevelocity of mercury at 25°C reported by Birch (1960).There is a small temperature dependence of ±0.2% fora range of 19° to 31°C which can be neglected. Goodsignals can be obtained only above 0.2 kbar pressureand the accuracy increases with increasing pressure asthe transducer-sample contact and thus signal im-proves. The mercury-delay-line reading for zero samplelength was determined by extrapolating the pulse travel

times of five precision machined steel cylinders 1 to 5cm in length. They also were used as internal standardsfor periodic calibration checks. The length of the sam-ple is measured with a precision caliper. The velocity inthe rock then is calculated from the length of the sam-ple, the length of the mercury column, and the velocityof the mercury. There are differences in velocity com-monly of up to 0.5% between increasing and decreasingpressure, but there appears not to be a systematicdifference. The reproducibility for measurement on asingle sample is about ±0.4% for compression and±0.8% for shear, and the values at different pressuresfor one measurement have a relative accuracy of about±0.2% for compression and 0.4% for shear. The ab-solute accuracies at 0.5 kbar and above are estimated tobe ± 1 % for compression and ±2% for shear. This es-timated error is about double that of shoremeasurements largely because of the difficulty in get-ting the sample ends flat and parallel with the ship-board equipment and because of the high level of elec-trical and mechanical noise onboard the ship. It still isadequate for most purposes.

Shore-laboratory measurements to 10 kbar made es-sentially with the same procedure as that describedabove for onboard ship are reported by Christensen(this volume). The accuracy is about ±0.5% for com-pression and ± 1 % for shear. The compressional veloci-ty of six samples was measured both onboard ship andin the shore laboratory. At 0.5 kbar the mean differenceof pairs of values is 1.4% with the shipboard velocitieshaving a mean value that is 0.8% higher. This issatisfactory agreement particularly as the sample faces

380


were remachined before the shore measurements. Shearvelocity measurements were repeated on only twosamples, the shipboard values averaging 3% higher.Thus, there may be a small but barely significantdifference in calibration for shear wave velocities in theshipboard and shore laboratories.

The usefulness of laboratory measurement of veloci-ty depends on how well the in situ conditions aresimulated, notably the pressure, temperature, extent ofwater saturation, and the sample orientation if it isanisotropic. There also is the serious problem of howrepresentative the samples are of the complete sectionor region being investigated. This latter problem is dis-cussed later. The velocities of the sea-floor rocks in-crease considerably but variably with increasing con-fining pressure. The measured velocities of Leg 37samples increase from atmospheric pressure, to 0.5kbar (upper crust) by about 0.2 km/sec, to 2.0 kbar(lower crust) by about 0.3 km/sec, and to 10.0 kbar byabout 0.5 km/sec. Older, more-weathered samplescommonly have increases twice those above, particular-ly at the lower pressures. The prime source of in-creasing velocity with pressure up to 2.0 kbar is theclosing of microcracks and pores. The confiningpressure, which is effective in closing such cracks andpores, is equal to the total external pressure minus theinternal or pore pressure. In the laboratory the porepressure is small and the confining pressure close to theexternal pressure because the pore water usually isallowed to escape as the pressure is increased, andbecause the compressibility of the pore water is muchgreater than that of the rock. The total or lithostaticpressure on an in situ rock is equal to the sum of theseawater load, 0.10 kbar per km depth which is 0.18kbar for Site 332 and the rock load, about 0.20 kbar perkm of sediment, and 0.28 kbar per km of basalt. Thetotal pressure at the bottom of the 722-meter deep Hole332B is about 0.37 kbar. The pore pressure withincrustal rocks may vary between close to zero andhydrostatic pressure where all pore spaces are con-tinuously connected with the seawater. If the pores andcracks are not connected, the pore pressure generallywill be small because the thermal contraction of thepore water is much greater than that of the rock in cool-ing from a high temperature of formation since thecoefficient of thermal expansion of water is about fourtimes that of basalt and because the compressibility ofthe water is much greater than that of the rock in goingfrom a shallow depth of formation to subsequent deepburial because the compressibility of water is about 20times that of basalt. Only in the special but perhaps im-portant situations of increasing temperature or of up-ward tectonic displacement will the pore pressure ofcrustal rocks exceed hydrostatic and approachlithostatic pressure so that the confining pressure issmall. Thus, the in situ confining pressure generally willrange from, (a) the difference between total load orlithostatic pressure and hydrostatic pressure, i.e., 0.12kbar at the bottom of Hole 332B to (b) lithostaticpressure, i.e., 0.37 kbar at the bottom of Hole 332B.

In addition to the pressure effect of the rock as it wason the sea floor, the drilling process and the stresses

from rapid reduction in pressure and temperaturechange at the time of sample recovery probably willcause extensive microcracks and fractures that willreduce the velocities measured at low pressures. Theywill be closed at about 0.5 kbar pressure. Thus, it isemphasized that laboratory pressure can be relatedonly roughly to equivalent depth beneath the sea floor.The 0.5-kbar velocities are taken to be representative ofupper crustal layer 2 and 2.0-kbar velocities to berepresentative of lower crustal layer 3. The errorresulting from incorrectly simulating in situ pressuresshould be less than 0.1 km/sec or 2% for both com-pressional and shear waves.

The temperature effect on velocity has not beenstudied in much detail but is of the order of -6 ×1 0-5oC- i (e g ? B i r c h i 1 9 5 8 ) f o r m a f i c r o c k s T m α S 5 t h e

difference between laboratory temperature of 25 ±5°Cand in situ temperature of about 5° to 100° (Hyndman,et al., this volume) in layer 2 is less than 0.5% and canbe neglected. In some regions of layer 2, and more fre-quently in layer 3 very near a spreading ridge wheretemperatures reach several hundred degrees, the errormay be several percent. The velocities probably will bequite different at temperatures approaching melting.

The importance of the measured samples being watersaturated has been emphasized by Dortman and Magid(1969), Nur and Simmons (1969), Christensen (1970),Christensen and Salisbury (1975) for representativecompressional velocities. The effect of saturation issmall for shear velocities. All of the samples were watersaturated before measurement by immersion inseawater following evacuation, then being held at 1.0-kbar pressure in seawater. As discussed in the sectionon the porosity of basalts (Hyndman and Drury, thisvolume), even pressure saturation does not completelyresaturate these rocks after thorough drying, but theeffect on velocity does not appear to be significant.Probably if the sample is nearly saturated, for example75% or more, the excess water produced with increasingpressure during measurement rapidly fills any unfilledpores and cracks. Also the larger cracks, which aremost easily filled, have the greatest effect on velocity.The error from incomplete saturation should be lessthan 0.5% for compressional velocities up to 1 kbar andnegligible above 1 kbar and for shear velocities.

The velocities all have been measured on transverseminicores 2.5 cm diameter × 3 to 7 cm length takenfrom most of the major cooling units recovered whichare generally equivalent to flows. The orientation ofwave propagation thus is horizontal with randomazimuth. Previous measurements have indicated anegligible velocity anisotropy in basalts (e.g.,Christensen and Shaw, 1970; Christensen andSalisbury, 1972).

VELOCITY RESULTSCompressional and shear wave velocities along with

sample densities and Poisson's ratios are given in Table1 with a few representative sample results being shownin Figure 1. The mean compressional velocity of basalt

381

R. D. HYNDMAN

6 . 2 r

3.0 -

0.5 1.0 1.5Pressure (Kbars)

2.0

Figure 1. Examples of compressional and shear wave velo-cities to 2.0-kbar confining pressure measured onboardGlomar Challenger. The solid points are for increasingand open circles decreasing pressure.

samples does not vary significantly among the fiveholes, and there is no systematic trend with depth. Themean and standard deviation of individual values of 79basalt samples, mainly Site 332, at 0.5 kbar correspond-ing to an upper crustal pressure, measured onboardship is 5.94 ±0.34 km/sec. The mean shear velocity of37 samples is 3.26 ±0.16 km/sec. The ratio of com-pressional to shear velocity is very constant at 1.86±0.02 giving a Poisson's ratio of 0.295 ±0.002. Themean bulk density of 101 basalt samples is 2.80 ±0.08g/cm3. Histograms of numbers of samples versus com-pressional and shear velocity and density are shown inHyndman and Drury, this volume.

The mean values and standard deviations for six gab-bros are: at 0.5 kbar, compressional velocity, 7.13±0.19 km/sec; shear velocity, 3.76 ±0.07 km/sec;Poisson's ratio 0.31 ±0.01; and density, 2.96 ±0.07g/cm3. At the lower crustal pressure of 2.0 kbar thecompressional velocity is 7.21 km/sec and shear veloci-ty is 3.79 km/sec.

The mean compressional velocity of three serpen-tinized peridotites at 2.0 kbar is 6.1 km/sec and for onesample the shear velocity was 2.17 km/sec giving aPoisson's ratio of 0.37. The mean density of threesamples is 2.71 g/cm3.

RELATIONS BETWEEN VELOCITY ANDDENSITY AND BETWEEN POISSON'S

RATIO AND VELOCITY

The variation of compressional and shear velocity ofbasalts has previously been shown to be closely andnearly linearly correlated with bulk density (e.g.,

Christensen and Salisbury, 1975) following the moregeneral empirical relation suggested by Birch (1961).This relation is very useful for relating refractionmeasurements of velocity and gravity estimates of den-sity. The relation for Leg 37 basalts, gabbros, andserpentinized peridotites is shown in Figure 2. Ourbasalt velocities are slightly higher or densities lowercompared to previous values, but the difference is notsignificant. Both the gabbros and serpentinizedperidotites have significantly higher velocity for a givendensity compared to the basalts, particularly for com-pressional velocity. The much higher ratio of com-pressional to shear velocities, and thus the Poissonratio, for the serpentinized peridotites compared to thebasalts is readily apparent. Note on the same diagramthe agreement between the small range of velocities anddensities for our basalts and previously measuredyoung unweathered samples with ages less than 20 m.y.The range including older, more-weathered basaltsamples is much larger.

The dependence of Poisson's ratio on compressionalvelocity for Leg 37 samples measured onboard ship isshown in Figure 3. There is a clear correlation ofdecreasing Poisson's ratio with decreasing velocity. Theregression line for the basalts alone is:

δ = 0.194 (±0.031) + 0.0167 (±0.0050) Vp

The gabbros appear to lie along the same line. Theobserved trend could be an artifact, the result of asystematic measurement error, particularly in the zero-sample-length mercury-delay-line calibration.However, the error required is about five times the es-timated error limits. Also, the comparison of shipboardand shore-laboratory measurements (see above) in-dicates that if an error exists it is in the shear velocitiesbeing too fast. An error in the zero-sample-lengthcalibration that gives shear velocities that are too fastwould produce a correlation of Poisson's ratio andvelocity the opposite from that observed. Thus, theabove correlation is accepted as valid.

Christensen (personal communication) has examinedthe correlation between Poisson's ratio and com-pressional velocity for DSDP basalts from many sitesof different basement ages. He found a wide scatter, butwith a general trend of increasing Poisson's ratio withdecreasing velocity, the opposite from that for Leg 37shipboard measurements. His values of Poisson's ratioare similar for high velocity samples, i.e., 6 km/sec, buttend to much larger values compared to those for Leg37 basalts for low velocity samples, i.e., 4 km/sec. Thedifferent correlation for the two sets of samples may ex-ist because there are different origins for the velocityvariation within each set. For Leg 37 basalts, which arevery young and little weathered, decrease in velocitycomes mainly from increasing vesicular porosity.Vesicular porosity probably affects the compressionalmore than shear wave velocity, i.e., mainly decreasesthe bulk modulus. In contrast, the samples examined byChristensen that are generally older probably havedecreasing velocity primarily associated with increasingweathering. Weathering affects the shear more than the

382


7.0

6.0

5.0

4.0

3.0

2.0

1.0

r

-

-

iRange of Vpfor 57 previousbasalts age < 20

Range for «al l previous ^basalts '

× ^

i i

mv Å\

/ Vp = -4

•

'** Vs = - 3

i

i.26

Λ

V

.07

1

o o θ o

+3.56C

^ '

+2.17 e

• basaltO gabbro

serp. perid.

i i2.0 2.2 2.4 2.6 2.8 3.0 3.2

Bulk Density (gm cm" )

Figure 2. Variations of compressional and shear wave velo-cities at 0.5-kbar as a function of bulk density for Leg 37basement samples. The relations for previous basalts fromChristensen and Salisbury (1975) is shown for compari-son. Note the good agreement between the small rangeof velocities for the Leg 37 basalts and previous youngunweathered basalts.

0.36

0.34

0.32

0.30

0.28

0.26

OSerpentinite

Basalts

i I I I I i i i I I I i i5.0 5.4 5.8 6.2 6.6 7.0 7.4

Compressional Velocity, Vp (km s-1

Figure 3. Relation between Poisson 's ratio (δ) and compres-sional velocity for Leg 37 basement samples. The least-squares regression line for the basalts also is shown.

compressional velocity, i.e., mainly decreases the shearmodulus, so increases Poisson's ratio (see Christensenand Salisbury, 1972, for examples).

The correlation between Poisson's ratio and com-pressional velocity for the young unweathered Leg 37

basalts is the one appropriate for comparison withrefraction measurements since weathering even in oldsea floor is limited largely to the upper 50 meters oflayer 2 (see above). The large-scale voids and fracturesthat are responsible for reducing the refractionvelocities in layer 2 below that for solid basalt (see dis-cussion below) have dimensions that are about thesame fraction of the wavelengths of about 50 metersused as have the vesicles of the laboratory wavelengthsof about 0.5 cm. If the assumption that the relationgiven above applies to layer 2 refraction measurementsis correct, the Poisson's ratio predicted for a layer 2avelocity of 2.8 km/sec is 0.24. This relation thus ex-plains why the Poisson's ratio for layer 2 by refraction(see compilation by Christensen and Salisbury, 1975)generally is less than the mean laboratory value of 0.30for basalts. The relation also suggests the value ofmeasuring shear velocity and thus Poisson's ratio inrefraction measurements to distinguish high velocitysediments, i.e., 3 to 4 km/sec, from a low velocitybasaltic layer 2. The sediments generally will havePoisson's ratios well above 0.30 and the low velocitylayer 2 section values well below 0.30.

CRUSTAL LAYERS 2a AND 2b

As discussed in the introduction, there is a seriousdiscrepancy between the low seismic refractionvelocities obtained for layer 2 of 5.0 ±0.7 km/sec andthe velocities measured on fresh sea-floor basalts. Thehigh laboratory velocities of young basalts is substan-tiated by our large collection of 79 basalts with a meanand standard deviation of individual values of 5.94±0.34 km/sec. The discrepancy is even larger in someregions such as the Reykjanes Ridge (Talwani et al.,1971) where the upper part of layer 2, designated layer2a, has velocities between 2.3 and 3.7 km/sec. Wheredetected, the layer has a thickness of about 1 km, but ifthe thickness was much less, the layer usually would notbe resolved. The layer also usually would not have beendetected where there is a significant sedimentarythickness because it then would not produce a firstarrival. A few very detailed seismic measurementssuggest that actually there may be a gradual rather thanstepwise increase in velocity with depth in the uppercrust (Sutton et al., 1971; Helmberger and Morris,1969; Hinz and Moe, 1971). Oceanic basementvelocities of less than 4.0 km/sec have been recorded ata number of locations on the Mid-Atlantic Ridge (LePichon et al., 1965; Keen and Tramontini, 1970).

R.B. Whitmarsh (personal communication) has ob-tained a reversed refraction line over the sites of thethree deepest holes of Leg 37, Sites 332 and 333, andfound a very low velocity upper basement layer. Thelayer has a velocity of 2.8 km/sec and is about 640meters thick (his model 1) underlain by 4.9 km/secmaterial. Most other refraction lines in the area do notshow a well-defined layer 2a and give normal layer 2velocities. A low velocity upper layer, however, doesappear to be present under the median valley (Fowlerand Matthews, 1974; Whitmarsh, 1973; Poehls, 1974).The highest layer 2 velocities in the Mid-Atlantic Ridgecrestal mountains region are about 5.7 km/sec (Le

383

R. D. HYNDMAN

Pichon et al., 1965). As discussed below, it is verysignificant that much more low velocity broken basaltand sediment was encountered in the upper basement atSites 332 and 333 than at the other sites.

Our drilling, particularly at Site 332, appears to haveresolved the discrepancy between laboratory velocitiesof basalts and layer 2 seismic refraction velocities. Theupper part of Hole 332B penetrated interlayered solidbasalt and broken and fractured basalt with some sedi-ment. The broken and shattered basalt and sedimentundoubtedly are responsible for the low refractionvelocities. Unfortunately, only small amounts of thebroken basalt and sediment were recovered in the core.The presence of the low refraction velocity surface layerunder the median valley and a comparison of the layerthickness with the observed sedimentation rates andavailable deposition time require that the low velocitymaterial be primarily shattered, porous basalt ratherthan sediment. The smooth variation with depth of thepaleomagnetic inclinations of samples, although withvery shallow inclinations (see Ade-Hall et al., thisvolume) imply that, at least the recovered material isnot talus but rather flow breccia, drained pillows andlava tubes, and generally fractured and shattered rockswith large voids, open fissures, or cracks such as fre-quently have been observed in the submersible studiesin the FAMOUS area of the median valley at 37°N(e.g., Heirtzler and Le Pichon, 1974). Lort andMatthews (1972) have shown that the fractured andshattered basalts of the Troodos ophiolite complexhave similarly low refraction velocities of about 3.0km/sec.

The high drilling water pressure required to clear thebit while drilling basalt tends to wash out any fracturedor unconsolidated material. However, the drilling rateis much higher and bit pressure much lower whenshattered basalt and sediment are being drilled. Bothdrilling rate and bit pressure are continuously recordedso the fraction of solid basalt and of shattered materialor sediment in a drilled section can be computed ap-proximately. We have assumed that the fraction ofsolid basalt in each section is proportional to aparameter we call the "relative drilling effort" (RDE).

Measured Velocity {%)

5.0 6.0 7.0 0 20 40 60

Estimated Velocity (km sec"1)

2.0 3.0 4.0 5.0 6.0

RDE = C×•R

where L is the bit load, R the drilling or penetrationrate, and C is a constant chosen so that the maximumRDE that commonly is observed equals 1. The relativedrilling effort and the computed fraction of solid basaltas a function of depth in Holes 332B and 333 are shownin Figure 4. Also the percent recovery is seen generallyto follow the drilling effort. To estimate the velocity ofthe composite material, we have used our meanlaboratory basalt velocity at 0.5-kbar pressure of 5.94km/sec and an estimate of mean rubble and sedimentvelocity of 2.2 km/sec, i.e., about 50% voids or lowvelocity sediment. This velocity, of course, is poorlydetermined. Wyllie et al. (1958) have shown that if thevelocities of two media are not too different, e.g., waterand rock, the velocity V through the composite

Figure 4. Compressional velocities with depth of basalts at0.5-kbar at Sites 332 and 333, the variation in relativedrilling effort, the computed ratio of solid basalt to bro-ken volcanic rubble and sediment, and the inferred velo-city-depth profiles. The refraction model for the sites byR. B. Whitmarsh is given for comparison.

material can be represented closely by the time averageequation:

_L=^_ 1 -A

where V and V2 are the velocities in the two materialsand A is the fraction of material with velocity V, and 1-A the fraction of material of velocity Vi.

The computed velocity-depth profiles for Holes 332Band 333A are compared with Whitmarsh's refractionmodel 1 for the site in Figure 4. Although our com-puted profiles show much more detail than the broadaverage layering obtained from the refractionmeasurements, there is general agreement in thethickness and velocity of the low velocity surface layer2a and in the velocity of the underlying layer 2b. Agradual increase in refraction velocity with depth in theoceanic basement, such as inferred from our drill holeshas been found wherever they have provided sufficientdetail. For example, Sutton et al. (1971) found thatvelocities of about 3.6, 4.5, and 5.8 km/sec occur fre-quently in layer 2. The computed profile implies up to35% voids or low velocity sediment near the top of thedrilled section decreasing to about 10% at a depth of700 meters.

MAFIC AND ULTRAMAFIC ROCKS ASPOSSIBLE CONSTITUENTS OF LAYER 3

Oceanic crustal layer 3 lies at an average depth of 1.7km beneath oceanic basement. It has a thickness of 4 to7 km and an average seismic refraction velocity of 6.7km/sec (Raitt, 1963; Ludwig et al., 1971; Christensenand Salisbury, 1975). The lower 3 km of layer 3, i.e., 3b,is frequently delineated as a separate "basal layer" with

384


refraction velocity of 6.7 km/sec where the seismic dataare sufficiently detailed (Sutton et al., 1971). Theaverage velocity of the upper part of layer 3, i.e., 3a,corresponds to the laboratory velocity of a number ofmafic and some ultramafic rocks. Since some lowvelocity material or fractures must be present (see Lortand Matthews, 1972) although much less than in layer2, laboratory velocities should be somewhat higherthan 6.7 km/sec. The smaller range of refractionvelocities for upper layer 3 compared to layer 2 un-doubtedly reflects a smaller fraction of low velocitymaterial. The most likely constituents of upper layer 3are gabbro, metagabbro, and metabasalt (Cann, 1968;Barrett and Aumento, 1970; Christensen, 1970; Fox etal., 1973; Christensen and Salisbury, 1975).Amphibolites can have the required velocity, but theyare not frequently dredged, and the high temperatureconditions for their formation probably do not occurfrequently in the oceanic crust. Serpentinites also canhave the required velocity, but are discounted except asa minor constituent because their range of velocity,depending on the degree of serpentinization, is muchgreater than the observed range of refraction velocitiesand because the Poisson's ratio, and ratio of com-pressional to shear velocity measured in the laboratory,are much higher than that obtained from seismic data(Christensen, 1966; Christensen, 1972; Christensen andSalisbury, 1975). Gabbros and low-grade metamorphicmafic rocks have the required velocities and Poisson'sratios and are frequently dredged.

The lower part of layer 3, the basal layer or 3b, hasseismic refraction velocities intermediate between thelaboratory velocities of gabbros and peridotite. Fewrocks have this velocity. A mix of gabbro and peridotiteis a possibility. Serpentinite is discounted except as aminor constituent as discussed above. Ophiolite com-plexes suggest a gabbroic upper layer 3 and a basallayer of mixed mafic and ultramafic rocks that fre-quently are layered cumulates. Serpentinites also areobserved, but much of the serpentinization probablyhas taken place after surface exposure of the ophiolites(Magaritz and Taylor, 1974).

Hole 334 penetrated first a 50-meter thick surfacelayer of basalts, then a complex sequence of gabbro,olivine gabbro, serpentinized peridotites, and brecciasconsisting of the latter three rock types with somesedimentary matrix. The hole was sited at the base of afaulted slope in an attempt to penetrate to deep crustalmaterial. The presence of the sedimentary brecciassuggests surface exposure of a plutonic melange priorto burial by subsequent basaltic eruptions. About 45%of the plutonic sequence is gabbro and olivine gabbro,25% serpentinized peridotite, and 30% breccia. There isno direct evidence that the drilled sequence representslower crustal material, but it is important to comparethe measured velocities of the rock types recovered withseismic refraction results. Olivine gabbros very fre-quently occur in the lower part of ophiolite complexesand may be characteristic of the lower crust, and thehigh density of the plutonic rocks (mean of gabbros,2.96 g/cm3, mean of serpentinized peridotites, 2.70g/cm3) argues against a diapiric intrusion and for anupward displacement by faulting. The more completelyserpentinized rocks commonly dredged have lower den-

sities, about 2.5 g/cm3, and could rise diapirically alongfault zones.

The compressional velocity of one serpentinizedperidotite at 2.0-kbar pressure corresponding to thelower crust, is 6.96 km/sec. The velocity is somewhathigher than that computed from the relation betweenvelocity and degree of serpentinization as determinedfrom the density. The density of 2.84 g/cm3 is ap-propriate for a velocity of about 6.0 km/sec accordingto the data of Christensen (1972). The sample musthave about 50% serpentine by volume. The shearvelocity is 3.21 km/sec, the ratio of compressional toshear velocity is 2.17, and Poisson's ratio is 0.36. ThePoisson's ratio is much higher than for basalts and gab-bros and much higher than found for layer 3 fromrefraction measurements (Christensen and Salisbury,1975). It agrees with the average serpentinite value of0.37 given by Christensen (1972).

The gabbros measured are very fresh with a meandensity of 2.96 g/cm3 and a water content from lowtemperature drying of only 0.4% by weight. They arecoarse grained with primary igneous textures. Thepyroxenes contain well-developed exsolution lamellaeof coexisting orthopyroxene and clinopyroxene. Themean compressional velocity and standard deviation ofsix individual values is 7.13 ±0.19 km/sec at 0.5 kbarincreasing to 7.21 ±0.19 km/sec at the lower crustalpressure of 2.0 kbar. The mean shear velocity at 0.5kbar is 3.76 ±0.07 km/sec and 3.79 ±0.10 km/sec at2.0 kbar. The ratio of compressional to shear velocity is1.89 and Poisson's ratio is 0.31, very slightly higherthan for the Leg 37 basalts.

The compressional velocity of 7.21 km/sec of thegabbros is close to 7.1 which is about the maximumcommonly recorded layer 3 (or 3a) velocity fromseismic refraction measurements. To obtain the averagelayer 3 velocity of 6.7 km/sec only 2% of cracks orother voids filled with water or other very low velocitymaterial is required if the travel time average equationof Wyllie et al. (1958) is correct. Alternatively about10% of material with velocity 4.5 km/sec is required.Faulting and fracturing of the crust certainly is suf-ficient for there to be enough voids and perhaps veinsof low velocity carbonates to produce this velocityreduction. The measured Poisson's ratio is slightlyhigher than the average but within the range fromrefraction, and as shown above, Poisson's ratio isreduced by the presence of voids and fractures.

The lower crustal layer 3b could consist of a mix ofgabbro and high velocity peridotite, e.g., our peridotitebefore serpentinization following surface exposure, asinferred from ophiolite sections (e.g., Moores andJackson, 1974; Peterson et al., 1974). About 25%peridotite with a velocity of 8.2 km/sec and 75% of ourgabbro would give the observed velocity of 7.4 km/sec.Less gabbro and more peridotite is required if fracturesand low velocity veins are important. Thus, the frac-tions we observe in our drilled section would give theobserved basal layer seismic refraction velocity.

CONCLUSIONSDeep drilling near the crest of the Mid-Atlantic

Ridge by Glomar Challenger has explained the lowvelocity (less than 4.0 km/sec) upper crustal layer 2a

385

R. D. HYNDMAN

outlined by seismic refraction measurements in someareas. In these areas, there is an upper layer of highlyfractured low density volcanic material, flow breccia,drained pillows, and lava tubes with extensive voidsand some intercalated sediments. The results also ex-plain why refraction velocities for layer 2 are in generalso much lower and more variable than the velocities ofsea-floor basalts measured in the laboratory. Therefraction velocity profile depends on the amount offracturing, voids, and sediment and not on a downwardchange in general rock composition or metamorphicgrade. The velocities measured in the laboratory onsolid basalt should be related to the maximum com-monly and accurately recorded refraction velocities.

Poisson's ratio for unweathered basalts is found todecrease with decreasing velocity. The relation predictsa Poisson's ratio of 0.28 for a layer 2 refraction com-pressional velocity of 5 km/sec and 0.24 for a layer 2avelocity of 2.8 km/sec in good agreement with thevalues observed.

Gabbros recovered from one hole on Leg 37 havevelocities appropriate to the upper part of crustal layer3. If the ultramafic rocks in Hole 334 had undergone lit-tle serpentinization, such a complex at the base of thecrust would explain the basal layer that has been ob-served by refraction measurements, with velocitiesbetween those of the oceanic layer 3 and the mantle.Similar mixed mafic-ultramafic complexes are commonat the base of the inferred crustal parts of ophiolites.

Much better recovery of the basement rocks is need-ed to determine the exact nature of the low velocitybroken material that makes up layer 2a. In-hole veloci-ty logging also now is required so that the velocity ofmaterial not recovered may be determined. Many morecrustal holes are needed to outline the extent and dis-tribution laterally of layer 2a. Deeper drilling, in excessof 2 km into the deep oceanic basement that is essentialfor the understanding of layer 3 now is clearly possiblewith the Glomar Challenger technology. We believe thata major effort should be made as soon as possible fordeep drilling into the oceanic basement.

ACKNOWLEDGMENTSI wish to express my appreciation for the opportunity to

participate in the Deep Sea Drilling Project. The officers andcrew of D/V Glomar Challenger and the DSDP technical staffonboard were helpful at all times. I thank Dr. Whitmarsh foruse of his unpublished seismic refraction results in the Leg 37drilling area, and Dr. N.I. Christensen for assistance with thedesign of the shipboard equipment and for many helpful com-ments on the analysis and the interpretation of the data. Thiswork was done while I was at the Department ofOceanography, Dalhousie University, Halifax, N.S., Canada.

REFERENCESBarrett, D.L. and Aumento, F., 1970. The mid-Atlantic Ridge

near 45°N. XI. Seismic velocity, density, and layering ofthe crust: Canadian J. Earth Sci., v. 7, p. 1117-1124.

Birch, F., 1958. Interpretation of the seismic structure of thecrust in the light of experimental studies of wave velocitiesin rocks: Contributions in geophysics in honor of BenoGutenberg: New York (Pergamon Press), p. 158-170.

, 1960. The velocity of compressional waves in rocksto 10 kb, Part 1: J. Geophys. Res., v. 65, p. 1083-1102.

., 1961. The velocity of compressional waves in rocks

Cann, J.R., 1968. Geological processes at mid-ocean ridgecrests: Geophys. J., v. 15, p. 331-341.

Cernock, P.J., 1970. Sound velocities in Gulf of Mexicosediments as related to physical properties and simulatedoverburden pressures: Texas A & M Univ. Tech. Rept. 70-5-T, Office of Naval Research, Contract N00014-68-A-0308 (0002 Oceanography).

Christensen, N.I., 1966. Shear wave velocities in metamorphicrocks at pressures to 10 kilobars: J. Geophys. Res., v. 71,p. 3549-3556.

, 1970. Composition and evolution of the oceaniccrust: Marine Geol., v. 8, p. 139-154.

., 1972. The abundance of serpentinites in the oceaniccrust: J. Geology, v. 80, p. 709-719.

Christensen, N.I. and Shaw, G.H., 1970. Elasticity of maficrocks from the mid-Atlantic ridge: Geophys. J., v. 20,p. 271-284.

Christensen, N.I. and Salisbury, M.H., 1972. Sea floorspreading, progressive alteration of layer 2 basalts, andassociated changes in seismic velocities: Earth Planet. Sci.Lett., v. 15, p. 367-375.

, 1973. Velocities, elastic moduli and weathering-agerelations for Pacific layer 2 basalts: Earth Planet. Sci.Lett., v. 19, p. 461-470.

., 1975. Structure and constitution of the lower

to 10 kb, Part 2: J. Geophys. Res., v. 66, p. 2199-2224.

oceanic crust: Rev. Geophys., v. 13, p. 57-86.Dortman, N.B. and Magid, M.S., 1969. New data on velocity

of elastic waves in crystalline rocks as a function ofmoisture: Internat. Geol. Rev., v. 11, p. 517-524.

Fowler, C.M.R. and Matthews, D.H., 1974. Seismic refrac-tion experiment on the mid-Atlantic ridge in the Famousarea: Nature, v. 249, p. 752-754.

Fox, P.J., Schreiber, E., and Peterson, J.J., 1973. The geologyof the oceanic crust: compressional wave velocities ofoceanic rocks: J. Geophys. Res., v. 78, p. 5155-5172.

Hamilton, E.L., 1965. Sound speed and related properties ofsediments from Experimental Mohole (Guadalupe Site):Geophysics, v. 30, p. 257-261.

Hart, R., 1973. A model for chemical exchange in the basalt-seawater system of oceanic layer II: Canadian J. EarthSci., v. 10, p. 799-816.

Heirtzler, J.R. and Le Pichon, X., 1974. A study of the genesisof tectonic plates: Geology, v. 2, p. 273.

Helmberger, D.V. and Morris, G.B., 1969. A travel time andamplitude interpretation of a marine refraction profile:primary waves: J. Geophys. Res., v. 74, p. 483-494.

Hinz, K. and Moe, A., 1971. Crustal structure in theNorwegian Sea: Nature, v. 232, p. 187-190.

Hyndman, R.D., 1974. Seismic velocities of basalts fromDSDP Leg 26. In Davies, T.A., Luyendyk, B.P., et al.,Initial Reports of the Deep Sea Drilling Project, Volume26: Washington (U.S. Government Printing Office),p. 509-512.

Keen, C.E. and Tramontini, C, 1970. A seismic refractionsurvey on the Mid-Atlantic Ridge: Geophys. J., v. 20,p. 473-491.

Le Pichon, X., Houtz, R.E., Drake, C.L., and Nafe, J.E.,1965. Crustal structure of the mid-ocean ridge, 1: seismicrefraction measurements: J. Geophys. Res., v. 70, p. 319-339.

Lort, J.M. and Matthews, D.A., 1972. Seismic velocitiesmeasured in rocks of the Troodos igneous complex:Geophys. J., v. 27, p. 383-392.

Ludwig, W.J., Nafe, J.E., and Drake, C.L., 1971. Seismicrefraction. In Maxwell, A.E., (Ed.), The sea, v. 4: NewYork (Interscience), p. 53-84.

Magaritz, M. and Taylor, H.P., 1974. Oxygen and hydrogenisotope studies of serpentinization in the Troodosophiolite complex, Cyprus: Earth Planet. Sci. Lett., v. 23,p. 8-14.

386


Moores, E.M. and Jackson, E.D., 1974. Ophiolites andoceanic crust: Nature, v. 250, p. 136-138.

Nur, A. and Simmons, G., 1969. The effect of saturation onvelocity in low porosity rocks: Earth Planet. Sci. Lett.,v. 7, p. 183-193.

Peterson, J.J., Fox, P.J., and Schreiber, E., 1974. New-foundland ophiolites and the geology of the oceanic layer:Nature, v. 247, p. 194-196.

Poehls, K.A., 1974. Seismic refraction on the mid-Atlanticridge at 37°N: J. Geophys. Res., v. 79, p. 3370-3373.

Raitt, R.W., 1963. The crustal rocks. In Hill, M.N., (Ed.),The sea, v. 3: New York (Wiley), p. 85-102.

Salisbury, M.H. and Christensen, N.I., 1976. Sonic velocitiesand densities of basalts from the Nazca plate, DSDP Leg34. In Hart, S.R., Yeats, R.S., et al., Initial Reports of theDeep Sea Drilling Project, Volume 34: Washington (U.S.Government Printing Office), p. 543-546.

Sutton, G.M., Maynard, G.L., and Hussong, D.M., 1971.Widespread occurrence of a high velocity basal layer in thePacific crust found with repetitive sources and sonobuoys.In Heacock, J.G., (Ed.), The structure and physicalproperties of the earth's crust, Geophys. Monogr. Ser.,v. 14: Washington (Am. Geophys. Union), p. 193-209.

Talwani, M., Windisch, G.C., and Langseth, M.G., Jr., 1971.Reykjanes Ridge crest: a detailed geophysical survey: J.Geophys. Res., v. 76, p. 473-517.

Whitmarsh, R.B., 1973. Median valley refraction line, Mid-Atlantic Ridge at 37°N: Nature, v. 246, p. 297-299.

Wyllie, M.R.J., Gregory, A.R., and Gardner, G.H.F., 1958.An experimental investigation of factors affecting elasticwave velocities in porous media: Geophysics, v. 23,p. 459-493.

387

of basements rocks from dsdp leg 37' r.d. hyndman

Documents