The use of submarine cables provides a tremendous opportunity for real-time data acquisition from permanent broadband seismometers on the seafloor. Programs to use retired submarine cables for this purpose have been initiated in the United States (e.g., Butler et al., 1995a) and Japan (e.g., Kasahara et al., 1998).

The Hawaii-2 submarine cable system is a retired American Telephone and Telegraph (AT&T) telephone cable system that originally connected San Luis Obispo, California, and Makaha, on Oahu, Hawaii (Fig. F1). The cable system was originally laid in 1964. In 1998, Incorporated Research Institutions for Seismology (IRIS) and scientists from the University of Hawaii and Woods Hole Oceanographic Institution installed a long-term seafloor observatory about halfway along the cable (~140W, 28N). The cable was cut and terminated with a seafloor junction box (Fig. F2). The location of the junction box on the seafloor defines the location of the Hawaii-2 Observatory (H2O), which was named after the original AT&T cable.

The junction box has eight underwater make-break connections. About 500 W of power is available from the junction box, and there is ample capacity for two-way, real-time communications with seafloor instruments. Data channels from the seafloor can be monitored continuously via the Oahu end of the cable to any laboratory in the world. The California end of the cable cannot be used because it was cut and removed from the continental shelf.

There is a shallow buried broadband seismometer operating at the site that monitored noise from the JOIDES Resolution during our cruise. The sensor consists of a modified Guralp CMG-3T broadband seismometer and a conventional 4.5-Hz three-component geophone, and it is buried in a caisson ~1 m below the seafloor (mbsf) (Duennebier et al., 2000, 2002). This sensor has been transmitting seismic data to shore continuously and in real time for >2 yr. The seismic data are forwarded to the IRIS Data Management Center in Seattle and are included in the Global Seismic Network database for use in global and regional earthquake studies. Other seafloor observatories, such as a geomagnetic observatory (Chave et al., 1995), a hydrothermal observatory (Davis et al., 1992; Foucher et al., 1995), or a broadband borehole seismic observatory (Orcutt and Stephen, 1993), can be installed at the site as funding becomes available.

Within the Ocean Drilling Program (ODP) and marine geology and geophysics communities, there has been considerable interest in the past few years in long-term seafloor observatories that include a borehole installation. Prototype long-term borehole and seafloor experiments almost exclusively use battery power and internal recording. The data are only available after a recovery cruise. One exception to this is the Columbia-Point Arena ocean-bottom seismic station (OBSS), which was deployed on an offshore cable by Sutton and others in the 1960s (Sutton et al., 1965; Sutton and Barstow, 1990). For the foreseeable future, the most practical method for acquiring real-time, continuous data from the seafloor will be over cables (Chave et al., 1990). The H2O project provides this opportunity.

Geological Setting

The Hawaii-2 cable runs south of the Moonless Mountains between the Murray and Molokai Fracture Zones (Fig. F1) (Mammerickx, 1989). Between 140and 143W, water depths along the cable track are typical for the deep ocean (4250-5000 m); the crustal age varies from 45 to 50 Ma (Eocene); and the sediment thickness to within the available resolution is ~100 m or less. Prior to the cable survey cruise in August 1997 (Stephen et al., 1997), sediment thickness was not well resolved along the track (Winterer, 1989).

Tectonically, the cable runs across the "disturbed zone" south of the Murray Fracture Zone, between magnetic isochrons 13 and 19 (Atwater, 1989; Atwater and Severinghaus, 1989). In the disturbed zone, substantial pieces of the Farallon plate were captured by the Pacific plate in three discrete ridge jumps and several propagating rifts. To avoid this tectonically complicated region and to be well away from the fracture zone south of the disturbed zone, the H2O was situated west of isochron 20 (45 Ma) at ~140W. The crust west of 140W was formed between the Pacific and Farallon plates under "normal" spreading conditions at a "fast" half-rate of ~71 mm/yr (Atwater, 1989; Cande and Kent, 1992). At the time this crust was formed, the Farallon plate had not split into the Cocos and Nazca plates, and the ridge that formed this crust was the same as the present-day East Pacific Rise.

Between 140 and 143W, the Hawaii-2 cable lies in the pelagic clay province of the North Pacific (Leinen, 1989). The sediments in this part of the Pacific are eolian in origin, consisting primarily of dust blown eastward from the arid regions of central Asia. This region of the Pacific is below the calcite compensation depth (~3500 m), and little or no biogenic calcite is thought to reach the seafloor (Leinen, 1989). Siliceous biogenic material is rapidly dissolved by the silica-poor bottom waters. At the seafloor, the sediments are unfossiliferous red clays.

The H2O site lies in a smooth abyssal plain environment. The drill site, identified as Site H2O-5 during planning and now identified as ODP Site 1224, is on the same crustal block as the H2O junction box (Table T1; Figs. F3, F4).

Scientific Objectives

Drilling at the H2O site was proposed to accomplish two main objectives:

  1. Drill and case a reentry hole into basement near the existing Hawaii-2 cable and the H2O junction box in order to establish a long-term borehole geophysical observatory for continuous real-time seismic monitoring, as well as other geophysical experiments.
  2. Sample a section of normal, fast-spreading ocean crust for use in constraining geochemical and hydrothermal models of crustal evolution.

Ocean Seismic Network

Establishing a borehole seismometer in the H2O area is valuable for addressing both teleseismic (whole Earth) and regional seismic studies. For uniform coverage of seismic stations on the surface of the planet, which is necessary for whole-Earth tomographic studies, seafloor seismic observatories are required. This site, where there is no land within a 1700-km radius, is one of three high-priority prototype observatories for the Ocean Seismic Network (OSN) (Butler, 1995a, 1995b; Purdy, 1995). Global seismic tomography (GST) provides three-dimensional images of the lateral heterogeneity in the mantle and is essential in addressing fundamental problems in subdisciplines of geodynamics such as mantle convection, mineral physics, long-wavelength gravimetry, geochemistry of ridge systems, geomagnetism, and geodesy. Specific problems include the characteristic spectrum of lateral heterogeneity as a function of depth, the anisotropy of the inner core, the structure of the core/mantle boundary, the role of oceanic plates and plumes in deep mantle circulation, and the source rupture processes of Southern Hemisphere earthquakes, which are among the world's largest (Forsyth et al., 1995).

The culturally important earthquakes in California are only observed at regional distances on land stations in North America, which restrict the azimuthal information to an arc spanning ~180. To observe California earthquakes at regional distances to the west requires seafloor stations. Regional observations are used in constraining earthquake source mechanisms. Since the H2O data will be available in real time, data could be incorporated into focal mechanism determinations within minutes of California earthquake events. Other problems that can be addressed with regional data from Californian and Hawaiian earthquakes are the structure of the 410-, 525-, and 670-km discontinuities in the northeastern Pacific and the variability of elastic and anelastic structure in the Pacific lithosphere from Po and So (Butler 1995a, 1995b).

In 1998 at the OSN pilot experiment site established in seafloor west of Hawaii, we deployed seafloor, buried, and borehole broadband seismometers to compare the performance of three different styles of installation. Figures F5 and F6 summarize for vertical and horizontal component data, respectively, the improvement that we expect to see in ambient seismic noise on placing a sensor in basement rather than on or in the sediments. Above the microseism peak at 0.3 Hz, the seafloor, buried, and borehole spectra at the OSN-1 site show the borehole to be 10 dB quieter on vertical components and 30 dB quieter on horizontal components (Collins et al., 2001). Shear wave resonances (or Scholte modes) are the physical mechanism responsible for the higher noise levels in or on the sediment. The resonance peaks are particularly distinct and strong at the H2O site. Note the 15-dB peak on the vertical component and the 35-dB peak on the horizontal components near 1 Hz on the H2O spectra. By placing a borehole seismometer in basement at the H2O site, we expect to eliminate these high ambient noise levels.

Basement Drilling on the Pacific Plate

In >30 yr of deep ocean drilling prior to ODP Leg 200 at more than 1200 sites worldwide, there have been only 13 holes with >10 m penetration into "normal" igneous Pacific plate: only one hole during ODP, only one hole with >100 m penetration, and no holes in crust with ages between 29 and 72 Ma. Table T2 summarizes the boreholes drilled on "normal" crust on the Pacific plate that have >10 m of basement penetration and crustal ages <100 Ma. Holes in seamounts, plateaus, aseismic ridges, and fracture zones were not included. Holes with crustal ages >100 Ma are not included because they would be affected by the mid-Cretaceous superplume (Pringle et al., 1993).

Besides the general sparsity of sampling of oceanic crust, there are no boreholes off-axis in "very fast" spreading crust. Although fast-spreading ridges represent only ~20% of the global ridge system, they produce more than one-half of the ocean crust on the surface of the planet, almost all of it along the East Pacific Rise. Most ocean crust currently being recycled back into the mantle at subduction zones was produced at a fast-spreading ridge. If we wish to understand the Wilson cycle in its most typical and geodynamically significant form, we need to examine ocean crust produced at fast-spreading ridges. We have also known for >40 yr that crust generated by fast spreading is both simple and uniform, certainly so in terms of seismic structure (Raitt, 1963; Menard, 1964). Successful deep drilling of such crust at any single location is thus likely to provide fundamental information that can be extrapolated to a significant fraction of the Earth's surface. Seafloor spreading that generated the ~45 Ma crust at the H2O was fast, with the full rate averaging 142 mm/yr. Thus, one objective of Leg 200 was to provide a reference station in "normal" fast-spreading ocean crust for use in constraining geochemical and hydrothermal models of crustal evolution.


A synopsis of the Leg 200 operations is given in Table T3. A summary of the time spent on various activities is given in Figure F7, and the operations summary is given in Table T1. The coring summary for Site 1224 is given in Table T4. We departed Site 1223 at 0130 hr on 23 December 2001 and arrived in the vicinity of the H2O junction box (2752.916N, 14159.504W) at 0000 hr on 26 December to begin a seismic and 3.5-kHz echo sounder survey.. All times are reported in ship local time, which is Universal Time Coordinated (UTC) - 9 hr at Site 1224. The 766-nmi voyage took 2.9 days at an average speed of 10.9 kt.

Following completion of the surveying at 0745 hr on 26 December, the JOIDES Resolution returned and positioned on proposed Site H2O-5 (Fig. F8) with Global Positioning System (GPS) navigation at 0845 hr on 26 December. Operations were suspended while waiting on weather (WOW) because of heave, pitch, roll, and wind up to 7.7 m, 5.2, 4.5, and 29 kt, respectively. A total of 13.25 hr of WOW time occurred before drilling operations could proceed.

Prior to conducting drilling operations, the vibration isolated television (VIT) camera was launched to conduct a camera survey of the site for debris, while also conducting a survey with an echo sounder attached to the VIT frame to further delineate subsurface layers. The survey covered a 30 m x 30 m area, took 2.0 hr, and showed the site was flat, undisturbed, and free of debris and cables.

First Jet-In Test

Drilling operations began when seafloor was tagged at 4966.1 meters below sea level (mbsl), or 4977 meters below rig floor (mbrf), at 1525 hr on 27 December. The jet-in test was performed to confirm a refusal depth for jetting in the reentry cone with 20-in casing. At 12 to 13 mbsf a hard layer was encountered, although the test was suspect as the ship was experiencing 4- to 5-m heave at the time of the test. Following the test, we were again forced to WOW, this time for 13.75 hr.

Hole 1224A

Hole 1224A was spudded at 1455 hr on 28 December at 4977 mbrf. Core 200-1224A-1X was advanced 6 m downhole with no recovery; hence, we could not establish a precise mudline (Table T4). On Core 200-1224A-4X, drilling progress was slow when we got to hard rock, which at the time was thought to be chert or basaltic basement. Recovery of 1.24 m of red clay and pieces of basalt confirmed that we had penetrated basement near the bottom of the 5.5-m interval cored or at ~28 mbsf. We attempted one more extended core barrel (XCB) core (200-1224A-5X), before switching to the motor-driven core barrel (MDCB) for one last short core. We pulled out of the hole and cleared the seafloor at 0530 hr on 29 December, ending Hole 1224A.

Overall we cored 32.2 m in Hole 1224A and recovered 1.67 m of core (5.19% recovery), with 32 m cored and 1.45 m recovered (4.53% recovery) with the XCB, and 0.2 m cored and 0.22 m recovered (110% recovery) with the MDCB (Table T4).

Hole 1224B

Hole 1224B was spudded with the advanced piston corer (APC) at 0650 hr on 29 December, but only 0.2 m of core was recovered. The primary goal of APC coring was to establish the mudline; therefore, we offset to spud Hole 1224C. Hole 1224B officially ended at 0745 hr on 29 December after we pulled the bit up to clear the seafloor.

Hole 1224C

The bit was positioned at 4964.1 mbsl (4975 mbrf), and Hole 1224C was spudded with the APC at 0820 hr on 29 December. We recovered 6.53 m of core and established the mudline at 4967.1 mbsl (4978.0 mbrf). Having successfully determined the mudline, the bit was pulled clear of the seafloor at 0915 hr on 29 December, marking the end of Hole 1224C.

Second Jet-In Test

Operations were delayed by bad weather, which included maximum heave, pitch, and roll of 6.3 m, 2.4, and 8.1, respectively, with winds up to 44 kt. Total time WOW was 16.0 hr, with operations beginning again at 0115 hr on 30 December.

A second jet-in test was deemed necessary to confirm the depth of penetration for the 20-in surface casing, which would be run with the reentry cone. A wash barrel was dropped, and the bottom-hole assembly (BHA) was jetted in to 4996.1 mbsl (5007 mbrf), ~29 mbsf, with no obstructions encountered, unlike the first jet-in test. The drill string was pulled out of the hole, with the bit clearing the rotary table at 1430 hr on 30 December.

Hole 1224D

The reentry cone was positioned over the moonpool doors, and the casing string was partially assembled at 1830 hr on 30 December 2001. Poor weather conditions and the associated large heave, roll, and pitch forced us to delay operations until 1715 hr on 1 January 2002, a loss of 46.75 hr.

With weather conditions improving, the reentry cone and ~25 m of 20-in casing were assembled and lowered through the moonpool at 2335 hr on 1 January. Hole 1224D was spudded at 1220 hr on 2 January. It took only 24 min to jet the 20-in casing string down to 5003.47 mbrf (25.47 mbsf) and set the reentry cone. VIT observation of the reentry cone confirmed that it was in a satisfactory position. The bit cleared the seafloor at 1315 hr on 2 January, and the pipe was tripped back to the rig floor, with the jet-in BHA and bit clearing the rotary table at 0200 hr on 3 January.

Hole 1224D was reentered with an RCB bit at 1837 hr on 3 January, with coring beginning at 25.5 mbsf. Coring progressed down to 59 mbsf, with several delays caused by the poor weather conditions, adding up to another 26.0 hr of WOW. The marine forecast was for continued poor weather for our operating area, with very strong low-pressure systems to the west and north and large swells. It was therefore decided to prepare to take advantage of any weather window by tripping the drill string and changing to the 14-in bit and BHA. This would allow us to open the cored hole when a more appropriate weather window was available and be in position to run 10-in casing. At 0045 hr on 7 January we started to trip the pipe, with the bit clearing the rotary at 1100 hr on 7 January.

Overall we cored 33.5 m in Hole 1224D and recovered 15.65 m of core (46.72% recovery) with the RCB coring system (Table T4).

Reaming the Hole

After tripping the pipe with the 14-in bit to 4790 mbrf at 2345 hr on 7 January, operations were again put on hold while WOW for 19.0 hr. Operations resumed at 1845 hr on 8 January after Hole 1224D was reentered.

We reamed the hole to 64.7 mbsf before drilling difficulties halted penetration. This was sufficiently close to the planned drilling depth of 67 mbsf, so we ceased drilling at 0730 hr on 11 January, for a total depth in Hole 1224D of 64.7 mbsf. When the drill string was pulled to the surface, to the surprise of all, the bit had been left in the hole, thus explaining the drilling difficulties. The bit appeared to have been sheared off.

Installation of 10-in Casing

Starting at 2030 hr on 11 January, the drill crew began assembling the 10-in casing string, which consisted of five joints of 10-in (40.5 lb/ft) casing. Hole 1224D was reentered by the casing string at 1336 hr on 12 January. We noted during reentry that the reentry cone and skirt had settled by ~1.7 m below the original mudline. The casing string was run down and landed with the base at 5036.47 mbrf (58.47 mbsf) on 1515 hr on 12 January. The casing was cemented with 18.8 bbl of 15.5 ppg Class G cement. The first attempt to release from the casing hanger failed and resulted in the 10-in casing hanger being pulled up above the reentry cone. The casing hanger was landed again in the 20-in casing hanger, and this time the 10-in hanger released at 1715 hr on 12 January. The pipe was tripped up, with the running tool clearing the rotary table at 0530 hr on 13 January.

The BHA was assembled with a RCB bit and run down to 4388.89 mbrf in preparation for coring in Hole 1224E. Before starting Hole 1224E, we reentered Hole 1224D to ensure that the casing and cement were properly installed.

Hole 1224E

The JOIDES Resolution was offset 15 m to the southwest, and Hole 1224E was spudded at 1840 hr on 13 January at 4978 mbrf. We washed down the first 8 m and then took two punch, or push, cores (200-1224E-1R and 2R), which were acquired by lowering the RCB bit through the soft sediments without rotating the bit. Both cores sustained substantial drilling disturbance, but we were able to recover 10.52 m of sediment core in a 19.2-m-long interval from 8.0 to 27.1 mbsf, where recovery was virtually absent in the other holes.

Coring penetrated from 27.1 to 36.7 mbsf for Core 200-1224E-3R. Basement was tagged at 27.7 mbsf during coring. Recovery consisted of basaltic basement underlying a 5-cm-thick piece of hyaloclastite, into which basalt glass and clay pieces had been incorporated. This likely is the top few centimeters of the basement. After completing coring on Core 200-1224E-3R, the bit was pulled up by one stand of drill pipe to connect another joint of pipe. This placed the bit above the sediment/basement contact. After making the connection, the driller was unable to reenter the basement hole. After 1 hr of attempting to find the hole by rotating the bit on bottom, a new hole (1224F) was started.

Overall we cored 28.7 m in Hole 1224E and recovered 14.91 m of core (51.95% recovery) with the RCB coring system (Table T4).

Hole 1224F

The start of Hole 1224F is somewhat of an anomaly in the ODP nomenclature, since the bit never pulled totally out of Hole 1224E, but it did pull out of the basement portion of Hole 1224E. The distance between Holes 1224E and 1224F is likely no more than ~1 m. In any case, we began penetrating basement again at 1630 hr on 14 January in Hole 1224F.

For all the bad weather we had previously endured, we were due a good spell. Thus, coring proceeded without interruption except for the occasional wiper trip and one trip to replace the knobby joints with drill pipe. During the latter trip, which started at 2315 hr on 17 January after recovery of Core 200-1224F-11R, the bit was inadvertently pulled above the basement/sediment contact. The driller worked the drill string up and down with rotation in an attempt to reenter Hole 1224F. Instead, Hole 1224E was reentered five times before finally the bit went back into Hole 1224F. RCB coring proceeded after washing ~11 m of soft fill from the bottom of hole. Cores continued to be cut at a rate of ~6-8 hr/core, which was roughly twice as fast as cores cut from near the top of the basement. No core was recovered in Core 200-1224F-16R. The bit deplugger was run to remove potential obstructions, but Core 200-1224F-17R also had no recovery. Owing to time limitation, coring in Hole 1224F ended and preparations for logging began.

Overall in Hole 1224F, we penetrated 174.5 m, cored 146.8 m, and recovered 37.7 m of core (25.68% recovery) with the RCB coring system (Table T4).


The bit was released in the bottom of the hole at 2320 hr on 19 January. The hole was then displaced with 75 bbl of sepiolite mud. A free-fall funnel (FFF) was launched at 0442 hr on 20 January to facilitate reentries into Hole 1224F during future scientific experiments.

At 0730 hr on 20 January, the triple combination (triple combo) tool was prepared to run downhole (see "Logging" in "Operations" in the "Site 1224" chapter). The tool reached 5152 mbrf, which is only 0.5 m off the bottom of the hole. The first logging run was completed, and the tool was through the rotary table at 1520 hr on 20 January. For each logging run, the base of the pipe was lowered to 49.9 mbsf initially. As each run was made uphole, the pipe was pulled up from 49.9 to 34.5 mbsf to increase the open-hole interval for logging.

The second logging run was with the Formation MicroScanner/dipole sonic imager (FMS/DSI) tool. Three passes of this string were run uphole at 275 m/hr from the bottom of the hole to the basement contact (27.7 mbsf). The second logging run was completed, and the tool cleared the rotary table at 0525 on 21 January.

We had planned to test the three-component well seismic tool (WST-3) if time and weather conditions permitted. While testing the tool and air gun, three problems were found: a faulty circuit in the blast hydrophone of the air gun, an air leak from the air gun, and the WST-3 telemetry worked intermittently. The experiment was thus terminated because there was insufficient time to attempt to fix the problems and complete the planned shooting program to the WST-3. The time constraint on the logging program was determined by the departure time required to make the San Diego port call. The WST was back through the rotary table at 0945 hr on 21 January.

The VIT was launched starting at 1030 hr 21 January to observe the FFF at the top of Hole 1224. A large hole was observed in the seafloor from circulating the cuttings out of the hole. As a result the top of the FFF was observed at ~4980.5 mbrf (2.5 mbsf) with the three buoys just below the mudline, secured to the FFF by a 5/32-in steel cable. The end of the casing on the FFF is estimated to be at 6.2 mbsf.

The open end of the drill pipe cleared the seafloor and FFF at 1238 hr on 21 January. The VIT was recovered at 1445 hr, and the BHA cleared the rotary table at 2355 hr on 21 January, completing activity at Site 1224.

Principal Results

Borehole Seismic Installation

The primary objective of the cruise was to prepare a borehole in basaltic crust for the installation of a broadband borehole seismometer that will be connected to the Hawaii-2 Observatory (H2O) for continuous, real-time data transmission to the University of Hawaii. From Hawaii the data will be made available to seismologists worldwide through the IRIS Data Management Center in Seattle.

Proposed Site H2O-5 (2753.363N, 14158.758W) (Fig. F8) was selected for the seismometer installation. This is 1.48 km northeast (a bearing of 056) of the H2O junction box location. The bearing was chosen so that regional earthquake events from the Island of Hawaii would be on the same great circle path to both the shallow buried seismometer at the junction box and to the borehole seismometer to be installed at Site 1224. The range was chosen as a compromise between being sufficiently far away to not disturb other experiments at the junction box but still close enough to conveniently run a cable from the borehole to the junction box. The bathymetry slopes smoothly downward ~6 m from the junction box to Site 1224, and the two sites appear to be on the same crustal block in a relatively flat abyssal plain environment.

Hole 1224D (2753.370N, 14158.753W; 4967 m water depth) has a reentry cone and 58.5 m of 10-in casing, which was cemented into a 30-m-thick well-consolidated massive basalt flow underlying 28 m of soft, red clay (Figs. F9, F10). After cementing, the hole was washed to a depth of 5036 mbrf (58 mbsf or 30 m into basaltic basement). On reentering for the wiper trip, we noticed that the cone had settled ~1.7 m into the sediment.

After setting the reentry cone and casing in Hole 1224D, we drilled a single-bit hole (1224F) to 174.5 mbsf to acquire sediment and basalt samples for shipboard and shore-based analysis as well as to run a logging program. Hole 1224F is <20 m southeast of Hole 1224D, and measurements in Hole 1224F can be used to infer the structure surrounding the seismometer hole. We dropped a FFF in Hole 1224F so that future borehole experiments using wireline reentry technology can be conducted (Figs. F11, F12). For example, this would be a good site to compare measurements in a sealed hole in basement (1224D) with measurements in an open hole in basement (1224F).


Sedimentary Section

Sediments were obtained from parts of Holes 1224A, 1224B, 1224C, and 1224E. The sediments consist mostly of abyssal clays of varying color. Occasional coarser horizons are present as are horizons with varying densities of microfossils, both siliceous (radiolarians and sponge spicules) and calcareous (coccoliths and discoasters) (Fig. F13).

Core recovery from Holes 1224A and 1224B was not significant enough to characterize the sediments. One significant discovery, however, was the recovery of light-colored noneffervescing granules and pebbles from a depth of between 6 and 15.6 mbsf from Hole 1224A. These were found to be zeolite deposits that were interpreted to have infilled burrows.

The total sediment depth at Site 1224 is 28 m. The top 6.53 m, as characterized by a single piston core in Hole 1224C, is massive brown clay that gradually changes color to very dark brown. Radiolarian spicules are present throughout the section, but they increase with depth and are common at the bottom of the unit. Sponge spicules are not found near the top of the section but are common below 4.50 m.

In Hole 1224E, we recovered 10.52 m of clay in the interval from 8.0 to 27.1 mbsf, in which two punch cores were collected with the RCB coring system by pushing through the sediment without rotating. The clay varies in color between dark brown, very dark brown, black, and dark yellowish brown. The high disturbance due to the punch coring process causes the colors to be streaked and mottled throughout the hole. Most color changes are gradual. Light-colored granules and pebbles are found in the top few centimeters of Core 200-1224E-1R (~8 mbsf). Like the burrows in Hole 1224A, they do not effervesce and are thought to be infilled burrows. These sediments also contain small manganese nodules. They are up to 2 mm in width and may be irregular or elongated in shape. Coccoliths and discoasters are present below 17.5 m.

We did not have a paleontologist on board, but Bob Goll and John Firth from ODP-TAMU (Texas A&M University) examined the radiolarians and calcareous nannofossils, respectively, postcruise. Paleontological analysis of the calcareous nannofossils indicates that essentially the whole sedimentary sequence was deposited within a few million years of the crustal age of ~46 Ma.

Hard Rock Section

The basalt stratigraphy at the site is summarized in Figure F14. In this figure, the depth for the top of each core, except for the topmost cores into basement, is taken as the top of the cored interval, as is the ODP convention. The cored interval is determined from the drill string length, which is entered into the ODP database. The topmost basement cores in all four holes, however, assume that the top of basaltic basement lies at a constant depth of 28 mbsf, which was our best estimate based on all drill holes and jet-in tests. For these cores only, the recovered basalt is placed below this fixed depth, rather than at the top of the cored intervals. The basement depth of 28 mbsf is probably uncertain by about a meter, and there may be some slight relief to the top of basalts as well. This approach avoids assigning basalt recovery to depths that actually are above the point where basement was touched by the drill string.

The lithostratigraphy of basalts at Site 1224 is divided into three units. Unit 3, the deepest unit, is intermixed pillows and flows of no more that a few meters thickness each. At least two, and probably more, eruptive events are represented. Overlying Unit 3, Unit 2 is a succession of thin flows and pillows. Chemical analyses of these rocks are very similar, indicating that they are one eruptive composition. Two thick lava flows from Unit 1 cap the underlying units. The lower of these has the same composition as the basalts of Unit 2. These two flows may have accumulated in a structural depression or pond that formed, probably by faulting, after eruption of the last thin flows or pillows of Unit 2.

Apart from the interiors of the massive flows, the lava sequence was pervasively altered under conditions with variable oxygen fugacity. Hydrous fluids carrying dissolved metals flowed through cracks, cavities, and fractures in the formation and deposited iron oxyhydroxides and sulfide minerals in this porosity structure. The fluids penetrated several centimeters into the rock adjacent to the fractures and impregnated microfractures with the same material that was deposited in the larger veins. Later, carbonate-saturated fluids coursed through the same fractures depositing calcite (Fig. F15). Except for the interiors of the two upper massive lavas, most of the rock was at least partially transformed to secondary minerals by this process. Eventually, enough calcite precipitated to cement the originally fragile iron oxyhydroxides and sulfide minerals. The calcite cementing is the principal reason why the coring of basalts above ~60 mbsf at Site 1224 was so successful.

It is too early to say how warm the fluids might have been, although calcite and aragonite are the ideal minerals to use for oxygen isotope determinations and to estimate temperature for the cementation portion of these processes. Iron oxyhydroxides are a principal component of hydrothermal sediments deposited on volcanically active ridge axes near, but not at, high-temperature vents. Elsewhere on the flanks of the East Pacific Rise, basalt coring has not been successful in crust as old as Miocene, largely because of the absence of calcite vein cement. Thus the carbonate-lined veins in the Eocene rocks at Site 1224 may be evidence for sustained fluid flow at low temperature and far off-axis. There is only a thin layer of sediment at Site 1224, insufficient to seal off fluids circulating in the crust. Interaction of those fluids with oxygenated bottom water may be why most of the section cored exhibits mainly oxidative alteration. The exception to this is the massive basalts at the top. In those, fractures may have been so few and widely spaced that fluid flow was restricted. The oxygen fugacity of the small quantity of fluids moving along them was consequently reduced by reaction with adjacent wall rocks, allowing pyrite to precipitate.


Thin section examination of volcanic basement at Site 1224 (Holes 1224A, 1224D, 1224E, and 1224F) evidenced a relatively homogeneous mineral paragenesis. The main phases are plagioclase, clinopyroxene, opaque minerals, and rare pigeonite; therefore, the rocks can be classified as tholeiitic basalts. Olivine is rare and only a few small iddingsitized euhedral to anhedral groundmass crystals were found. Iddingsite is a typical alteration of olivine and is made up by a mixture of goethite and layer silicates (e.g., smectite). The majority of the basalt is holocrystalline (almost 100% crystals) to hypocrystalline (glass concentration <50%) and can be ascribed to lava flows. With increasing depth of coring, hypohyaline textures and volcanic glass contents >90% become common and indicate the presence of pillow fragments with chilled margins. The deepest samples recovered (~153 mbsf) also show textural features of holocrystalline massive lava flows. With regard to their granularity, the basalts range from aphanitic (difficult to distinguish the crystals in the groundmass with the naked eye) to aphyric (absence of phenocrysts), though rare plagioclase or plagioclase-clinopyroxene sparsely phyric basalts (phenocryst content <2%) have been also found. The relative size of the crystals in the groundmass is equigranular, and their distribution is isotropic. The groundmass is hypidiomorphic with the presence of euhedral- to anhedral-shaped crystals. The texture of the massive lava flow basalts is intergranular (with clinopyroxene in interstitial relationships with plagioclase) to subophitic (with plagioclase laths partially enclosed in clinopyroxene) and, more rarely, intersertal (with microcrystalline to glassy material between plagioclase). Hyalopilitic (with plagioclase laths and clinopyroxene crystals in a glassy matrix) to, more rarely, intersertal textures have been found in the pillow lavas. The grain size of the groundmass ranges from very fine grained (0.001-0.5 mm) to fine grained (0.5-1 mm).


Inductively coupled plasma-atomic emission spectroscopy (ICP-AES) data for K2O, TiO2, MgO, Ba, and Zr were obtained on samples from Site 1224. These were supplemented in the site report by shore-based X-ray fluorescence data kindly provided by S. Haraguchi. The rocks recovered at the top of the hole (28-62.7 mbsf) consist of massive fresh basalt, with only widely spaced and narrow veins containing carbonate minerals, clays, and pyrite. From 62.7 to 133.5 mbsf, the basalts are thin sheet flows and pillows. From 133.5 to 161.7 mbsf, they are somewhat more massive pillows and flows.

The basalts are differentiated normal mid-ocean-ridge basalt (N-MORB) with 2-3.5 wt% TiO2. The samples selected for analysis from Hole 1224D are scarcely altered, with loss on ignition (LOI) values ranging from 0 to 0.45 wt%. Concentrations of K2O (0.11-0.27 wt%) may be slightly elevated in three of ten samples analyzed from this hole, but most values and all those for Ba (9 to 18 ppm) are consistently lower than in many comparably differentiated MORB glasses from the East Pacific Rise. This may indicate a greater-than-average depletion of the mantle sources of basalts obtained at depths shallower than 60 mbsf from Hole 1224D. Alternatively, the rocks may have experienced a slight nonoxidative alteration in which these components were partially removed from the rock. This seems unlikely, however, given the more extensive oxidative alteration observed in rocks from Holes 1224E and 1224F, obtained only a few meters away.

Both TiO2 contents and Zr concentrations are determined precisely enough, and they are so little affected by alteration that they can be used to define a chemical stratigraphy (Fig. F16). Most of the basalts from Hole 1224D belong to one chemically uniform, extensively differentiated basalt flow >20 m thick. This overlies a second flow that is not quite so differentiated. This in turn overlies pillows that are geochemically similar. Below 133.5 mbsf, strongly differentiated ferrobasalts were recovered.

X-ray diffraction (XRD) analysis was carried out on one clayey pebble from the sediment and twenty-five vein materials within the basalt. Five distinct vein types were documented by XRD analysis: clay, carbonate, zeolite, quartz, and calcite/smectite (Fig. F17). Many vein minerals in the basement at Site 1224 are stable at low temperature and pressure (i.e., zeolite). Phillipsite, the principal zeolite present at Site 1224, is a low-temperature member of the zeolite group (Miyashiro, 1973). Smectite is also commonly found as a product of the alteration of volcanic ash and rocks from the seafloor and is present in most of the low-grade metamorphic terranes in the world.

Four of the vein types observed at Site 1224, clay (smectite-illite), carbonate (calcite-aragonite), quartz, and zeolite, are similar to veins observed at Sites 896 and 504 near the Costa Rica Rift (Alt, Kinoshita, Stokking, et al., 1993). These minerals are present in relatively lower temperature hydrothermal assemblages (probably <100C) at these sites (Laverne et al., 1996). Truly high-temperature vein assemblages, such as the actinolite and epidote veins found below 2000 mbsf at Site 504, were not found at Site 1224. The mineral laumontite in the illite veins indicates a high-temperature zeolite facies assemblage (Miyashiro, 1973). Aragonite generally forms at a higher temperature than calcite. These minerals indicate the local influence of warm hydrothermal fluids.


We used progressive alternating-field (AF) demagnetization of archive-half sections, one whole-core section, one working-half section, and discrete samples to characterize the paleomagnetic signal and resolve the magnetization components recorded in the recovered core. An unambiguous magnetostratigraphy could not be obtained from the only undisturbed core (Core 200-1224C-1H) that was recovered in the sedimentary section; the other sediment cores were extremely disturbed by drilling. In addition, we only had time for a cursory interpretation of the magnetization of the basaltic units, although fairly detailed demagnetization experiments were conducted on split cores and discrete samples.

The magnetization of the basalts should provide a valuable paleolatitude estimate for the Pacific plate at ~45 Ma. This age corresponds to the Pacific plate's abrupt change in motion relative to the hotspots as marked by the kink in the Hawaiian-Emperor hotspot track. A cusp in the Pacific plate apparent polar wander path (APWP) may also occur at this age, marking a change in the motion of the Pacific plate relative to the spin axis. The Pacific APWP and hotspot tracks together provide key constraints on estimates of the size of motions between hotspots, ultimately extending our understanding of mantle dynamics (Acton and Gordon, 1994). Additionally, the age also lies within the period (39-57 Ma) when the Hawaiian hotspot has been shown to have moved rapidly southward relative to the spin axis (Petronotis et al., 1994). If geomagnetic secular variation has been averaged by the basalt units and if secondary overprints caused by alteration do not mask the primary magnetization, then we should be able to obtain an accurate paleolatitude. Finally, rock magnetic studies of the basalts should help refine our understanding of the magnetization of the upper oceanic crust and its role in generating lineated marine magnetic anomalies.


Samples of different sediment types and from basaltic rock were collected at Site 1224 for aerobic and anaerobic cultivation, for deoxyribonucleic acid (DNA) extraction and analysis, for phylogenetic characterization, for total cell counts, and for determination of the live/dead ratio of indigenous microbial communities. Sediment suspensions and ground basalt material were used under oxygen-depleted conditions in the anaerobic chamber for the establishment of enrichment cultures. Aerobic cultivation was conducted using both seawater-based media and commercial Zobell's medium. Anaerobic cultures were based on reduced mineral media.

To evaluate the microbial background at Site 1224, ambient seawater samples were collected at 1 m below sea surface upwind of the JOIDES Resolution. The microscopically enumerated total cell counts in the surface water at Site 1224 were 1.4 x 104 cells/mL.

Sediment samples from Holes 1224C, 1224D, and 1224E were obtained from different depths ranging from the near-surface layer down to 24.9 mbsf. Bacteria were present in all sediment samples taken to 24.9 mbsf.

The amount of active bacteria was assessed in two representative sediment samples taken from the near-surface layer (Sample 200-1224C-1H-1, 0-5 cm) and from a depth of 25 mbsf (Sample 200-1224E-2R-5, 143-150 cm). As indicated by fluorescent signals after hybridization with the Bacteria-specific probe EUB338, the amount of metabolically active bacteria ranged in these sediment layers from 62% to 41% of the total cell counts, respectively (Fig. F18).

Microscopic investigation of thin sections revealed the first survey of the presence of eukaryotic microorganisms within the basement of the North Pacific Ocean, which are counted among the kingdom of fungi. Hyphae have been found within cavities, small fractures, and veins filled with CaCO3. These fungal structures were viewed by transmitted light microscopy, in which they appear with a brownish tinge. The net of fungal hyphae shown in Figure F19 filled the complete space spanning from the basalt/calcite boundary to the center of the cavity. The cross-sectional dimension of the hyphal network is 5-10 m with a length ranging from 50 to several hundred micrometers. The hyphae are typically interrupted at irregular intervals by cross-walls, so-called septa, which divide the entire fungal hyphae into single distinctive cells. Our results provide strong evidence for eukaryotic life in addition to bacteria in deep subsurface environments.

Physical Properties

In Hole 1224A, P-wave velocities of aphyric basalt from Cores 200-1224A-5X and 6N are ~5900 m/s and ~5800 m/s, respectively.

In Hole 1224C, the gamma ray attenuation (GRA) densities of sediments gradually decrease with increasing depth between 0 and 6.4 mbsf, corresponding to a color change from light brown to dark brown. Similarly we observed an unusual trend for bulk and dry densities in Hole 1224C, which decreases from ~1.52 to ~1.36 g/cm3 and from ~0.8 to 0.54 g/cm3, respectively. Porosities in Hole 1224C gradually increase from 71% to 80%. P-wave velocities from the P-wave logger (PWL), however, show a small increase from 1460 to 1500 m/s with depth between 0 and 6.4 mbsf. P-wave velocities from PWS3 contact probe measurements from Core 200-1224C-1H to 4H (between 0 and ~5.70 mbsf) range from 1525 to 1535 m/s. The P-wave velocity in Core 200-1224C-5H is ~1555 m/s, which is greater than other sections. Grain densities in Hole 1224C show a small increase from 2.782 to 2.831 g/cm3 for depths shallower than 2 mbsf. Between 2 and 6 mbsf, grain densities remain fairly constant between ~2.70 and ~2.74 g/cm3.

In Hole 1224D, bulk and dry densities increase from 2.7 to 2.9 g/cm3 and 2.6 to 2.8 g/cm3, respectively, in Core 200-1224D-2R. In Core 200-1224D-3R, bulk and dry densities decrease from 2.9 to 2.8 g/cm3 and from 2.8 to 2.7 g/cm3, respectively. In Cores 200-1224D-4R and 5R, they also decrease from 2.85 to 2.80 g/cm3 and from 2.8 to 2.7 g/cm3, respectively. Porosities remain at low values ranging from 4% to 9%. PWS velocities range from 4200 to 6500 m/s. Compressional wave velocity anisotropies for each sample are ~2%-10%. PWS velocities have a sinusoidal depth variation. They decrease between 25 and 35 mbsf, increase between 35 and 45 mbsf, and decrease again between 45 and 55 mbsf. This sinusoidal depth variation is also identified for Hole 1224F.

Between 25 and 60 mbsf, PWS velocities in Holes 1224E and 1224F have a similar trend to Hole 1224D. PWS velocities have a strong depth dependence. Compressional velocities separate into seven depth zones (Fig. F20):

  1. 30-38 mbsf: 5500-6000 m/s,
  2. 38-41 mbsf: 4200-5500 m/s,
  3. 41-61 mbsf: 5000-6000 m/s,
  4. 61-100 mbsf: 4500-5000 m/s,
  5. 100-138 mbsf: 4700-6000 m/s,
  6. 138-147 mbsf: 4000-4700 m/s, and
  7. >147 mbsf: 5500 m/s.

Zones 1-3 may be characterized as rather uniform basalt flow zones with a thin low-velocity (fractured) layer. Zone 4 is characterized as a slightly low velocity zone. Velocities of zone 5 are higher than those for zones 4 and 6. Zone 6 is highly fractured, characterized by the lowest velocities. Zone 7 corresponds to more uniform basalt layers.

P-wave velocities are scattered with increasing bulk density. P-wave velocity vs. porosity, however, has a good inverse correlation, as P-wave velocity decreases with increasing porosity. These two relations imply that P-wave velocities are not controlled by bulk densities, but are well controlled by porosities. Large porosities are associated with more fractured zones. If this is true, zones 2 and 6 are intensively fractured.

Logging and Correlation to Other Data

Based on shipboard preliminary log analysis at this site during Leg 200, we conclude that basement in Hole 1224F consists of at least five distinctive units (Figs. F21, F22):

Sediments: 0-28 mbsf,
Unit I: 28-45 mbsf,
Unit II: 45-63 mbsf,
Unit III: 63-103 mbsf,
Unit IV: 103-142 mbsf, and
Unit V: Deeper than 142 mbsf.

These layered formations can be distinguished using the continuous electrical resistivity, density, sonic, neutron porosity, magnetic field, and possibly spectral gamma ray logs. The existence of a conduit or large-scale fracture between 138 and 142 mbsf was detected by all the log tools including the temperature tool. In addition, the temperature tool reveals that the "warmer" fluid had a temperature of 4.6C at the time of the logging. The vicinity of the 138- to 162-mbsf zone is much more highly altered than other rocks penetrated by the hole, as indicated by the gamma ray logs. Because of the relative position of the tools located in the tool strings, some tools can resolve the top logged intervals like gamma ray, porosity, density, and sonic logs. On the other hand, the resistivity tools and FMS placed at the bottom of the tool string can resolve the formation properties near the bottom of the hole. The values of the magnetic fields calculated from the three-component inclinometer tool are invalid near the bottom of the pipe (~35 mbsf). In the logged intervals where all the tools overlapped, they provide consistent information to support the layered structural units based on these geophysical properties.

Core lithology, physical properties, well logs, and seismic reflection data from the site were compared. The upper sediment unit (0-28 mbsf) is a brown clay layer with radiolarians at shallow depth. The mean velocities by physical properties measurements are ~1500 m/s. Logging Units I and II, between 28 and 63 mbsf, are two massive basalt flows with fractures at roughly 45 mbsf. These two logging units combined thus correspond to lithologic Unit 1. The compressional wave velocities in logging Units I and II based on core measurements are ~5500 m/s. Smecite veins were found in these units. Logging Unit III (63-103 mbsf) is characterized by fractured basalt layers both in core recovered and in data collected by the FMS/DSI logging tool. Calcite veins were found in this unit. The compressional velocity is ~5000 m/s. Logging Unit IV (103-142 mbsf) is characterized by stacks of small pieces of pillow lavas. This layer has compressional velocities slightly higher than 5.0 km/s as measured on discrete core samples. Logging data, however, indicate that this unit is highly porous. Logging Unit IV also contains smectite veins. At the base of logging Unit IV, large variations are present on the caliper log, resistivity log, compressional and shear velocity logs, U and Th content, and the temperature log. Physical properties measurements also indicate that this unit is highly fractured. The presence of high U and Th contents suggests that this unit is a highly altered zone. Logging Units III and IV combined correspond to lithologic Unit 2. In logging Unit V, below ~142 mbsf, basalt sheet flows were found. This logging unit corresponds to lithologic Unit 3. The single-channel seismic (SCS) data suggest that this is the top of a massive basalt unit that extends deeper than our deepest drilling depth (174.5 mbsf). In comparing the above units with the SCS records, these unit boundaries extend many kilometers away from the site. With further analysis it should be possible to understand the nature of oceanic Layers 2A and 2B and their relationship to lithologic boundaries in ~45-Ma fast-spreading oceanic crust.

3.5-kHz Deep-Source Experiment

A long-standing problem in the red clay province of the eastern Pacific Ocean is to adequately resolve chert layers and basement in the presence of sediments <50 m thick. By lowering a battery-powered, free-running 3.5-kHz pinger to the seafloor on the VIT sled and recording the pulse on the ship's 3.5-kHz acquisition system, we hoped to increase the sound level incident on the seafloor, to improve the penetration into the subbottom, to reduce the footprint of the sound on the seafloor, and to increase the received signal levels. The deep-source 3.5-kHz experiment was carried out whenever the VIT camera was lowered to the seafloor either for reconnaissance surveys or reentries.

Examination of the deep-source 3.5-kHz records shows two prominent reflections at 13 and 38 ms below the seafloor (Fig. F23). Depending on the sound velocity in the seabed, these reflectors would be 10 to 13 m and 28 to 38 m deep. The continuity of these reflectors varies with time throughout the survey, whereas the ship moves only a few meters.

Our preliminary interpretation had been that the 13-ms reflection is present at an intermittent chert layer. The first jet-in test stopped abruptly at 13 m. Although chert layers within the sediments have been encountered at other drill sites in the eastern Pacific, nowhere at Site 1224 did we sample chert. The 13-ms reflection may correlate with a radiolarian-rich layer that was cored. Basalt cores were regularly acquired at 28-30 m depth, corresponding to the 38-ms reflector.

In summary, the deep-source 3.5-kHz experiment identified a previously unrecorded reflector at 38 ms below the seafloor that corresponded to basaltic basement. This reflector was not observed in the traditional 3.5-kHz survey conducted in 1997 or in the shipboard 3.5-kHz survey acquired while we came on site (Fig. F4). The 38-ms reflector, however, was observed beneath the H2O junction box.

Broadband Seismic Observations during the Leg

Drilling at the H2O provides a unique opportunity to observe drilling related noise from the JOIDES Resolution on a seafloor seismometer in the 0.1- to 80-Hz frequency band. The University of Hawaii operates a Guralp CMG-3T three-component broadband seafloor seismometer and a conventional three-axis geophone at the H2O. Data are acquired continuously and are made available to scientists worldwide through the IRIS Data Management Center in Seattle. During the cruise, Jim Jolly and Fred Duennebier at the University of Hawaii relayed sample data files to the JOIDES Resolution by file transfer protocol (FTP) over the shipboard telephone. We were then able to process data and study correlations with on-site activities and weather. The University of Hawaii also maintained a Web site showing H2O seismic data collected during the cruise (www.soest.Hawaii.edu/H2O/).

Seismic activity could be associated with wind speed, sea state, shear resonance effects in the sediments, whales, water gun shooting, earthquakes, passing ships, and drilling-related activities such as bit noise and running pipe (Fig. F24).


  1. Site 1224 (2753.367N, 14158.755W) was selected for the seismometer installation (Fig. F8). This is 1.48 km northeast (a bearing of 056) of the H2O junction box location. Hole 1224D (2753.370N, 14158.753W; 4967 m water depth) has a reentry cone and 58.5 m of 10-in casing that was cemented into 30 m of well-consolidated, massive basalt underlying 28-30 m of soft, red clay.
  2. A single-bit hole (1224F) was drilled to 174.5 mbsf, and we sampled a 146.5-m-thick section of basaltic oceanic crust. Hole 1224F is <20 m to the southeast of Hole 1224D. Cores and well logs were acquired to characterize the site. The physical properties measured from well logs indicate that the basaltic basement can be divided into five distinct units at 28-45 mbsf, 45-63 mbsf, 63-103 mbsf, 103-142 mbsf, and below 142 mbsf. This hole was equipped with a FFF so that it also could potentially be used for long-term borehole seismic experiments in the future.
  3. A suite of experiments and observations was conducted to investigate the role of microbial life in the deep biosphere. As a general trend, bacterial population numbers decreased with increasing depth, though the amount of metabolically active bacteria remained remarkably high at 41% to 62% of the total cell counts. Both the high total cell counts as well as the amount of bacteria with apparent physiological potential within the sediment layers suggest a higher contribution of sediment bacteria than has been previously assumed. The successful cultivation of lithotrophic oxidizing bacteria and the microscopic indication of further microbial structures within a cavity of basaltic rock confirm the presence and even activity of microbial life not only in deep marine sediments, but also in the Paleogene oceanic crust from the North Pacific.
  4. We tested a deep 3.5-kHz source that could be deployed on the VIT frame to inspect the shallow structure of the seafloor at a higher spatial resolution than conventional echo sounding.
  5. The total sediment depth at Site 1224 is 28 m. Paleontological analysis of the calcareous nannofossils indicates that essentially the whole sedimentary sequence was deposited within a few million years of the crustal age of ~46 Ma.
  6. The basalts are differentiated N-MORB with 2-3.5 wt% TiO2.
  7. The lithostratigraphy of basalts at Site 1224 breaks down in a straightforward manner into three categories of extrusive basalt. The deepest rocks cored (Unit 3) are intermixed pillows and flows of no more than a few meters thickness each. At least two, and probably more, eruptive events are represented. These were followed by a succession of thin flows and pillows (Unit 2). Chemical analyses of these rocks are very similar, indicating that they are one eruptive composition. The whole sequence is capped by two thick lava flows (Unit 1). The lower of these has the same composition as the basalts of Unit 2. These two flows may have accumulated in a structural depression or pond that probably formed by faulting after eruption of the last thin flows or pillows of Unit 2.
  8. Apart from the interiors of the massive flows, the lava sequence was pervasively altered under oxidizing conditions. Hydrous fluids carrying dissolved metals flowed through cracks, cavities, and fractures in the formation and deposited iron oxyhydroxides in this porosity structure. The fluids penetrated several centimeters into rock adjacent to fractures, impregnating microfractures with the same material that was deposited in the larger veins. Later, carbonate-saturated fluids coursed through the same fractures and additional new ones, depositing calcite. Except for the interiors of the two upper massive ponded lavas, most of the rock was at least partially transformed into secondary minerals by this process. Eventually, enough calcite precipitated to cement the originally fragile iron oxyhydroxides, and this is the principal reason why the coring of basalts at Site 1224 was successful.