IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.329.103.2011

Igneous lithostratigraphy, petrology, alteration,
and structural geology

In Hole U1365E, basement was cored from 71 to 124 mbsf (0–53 meters subbasement, of which 39.66 m was recovered [74.6% recovery]).

The recovered basement consists of sparsely to highly phyric massive basalt sheet flows and aphyric to sparsely phyric thin sheet flows. These lithologies were divided into 17 basement units based on changes in lava morphology, flow boundaries, texture, and phenocryst occurrence. The distribution of lithologic units is summarized in Figure F15, and further detail regarding the definition of igneous units may be found in “Lithostratigraphy, igneous petrology, alteration, and structural geology” in the “Methods” chapter (Expedition 329 Scientists, 2011).

Massive lava flow units (1–9, 12, 13, and 17) were divided according to changes in phenocryst abundance, grain size, and flow margins. Thin flow units (10, 11, and 14–16) were divided based on the presence of chilled margins and variations in phenocryst mineralogy. Breccia fragments were recovered in intervals 329-U1365E-5R-4, 17–57 and 57–53 cm, and 8R-3, 130–144 cm, and are interpreted to represent interflow contacts. These breccias are grouped into subunits within each flow unit (Subunits 5e, 6a, and 17a). In addition, a number of unusual opaque calcite samples termed “black calcite” were recovered at two flow boundaries. The black calcite samples are interpreted to be part of interflow brecciation, possibly part of a cavity, and are identified as subunits accordingly (Subunits 4d and 7c). Incipient brecciation and vein nets were recovered at a number of intervals; however, these were not classified into units or subunits. A small fragment of basalt was recovered in the core catcher of Core 329-U1365A-26H; however, it is heavily altered to saponite and iron oxyhydroxides and its position relative to the units in Hole U1365E is not certain. It was therefore not assigned to a unit.

Pillow basalt was not observed at Site U1365 because none of the morphologic indicators normally associated with pillow basalts (glassy rinds, concentric cooling fractures, vesicle patterns, or extended chill zones) were observed. One lava flow sample exhibits ropy pahoehoe texture (Sample 329-U1365E-6R-1, 5–14 cm; Fig. F16). The lack of other flow margins and the presence of massive units above and below this margin indicate that it belongs to a flow surface that came in direct contact with seawater.

Lithologic units

Units 1–9, 12, 13, and 17 (massive sheet flow)

Massive sheet flows occupy 33.2 m (85%) of the recovered core and are the most abundant lava morphology at Site U1365. Classification of massive sheet flows was based on the presence of continuous sections of the same lithology and coarsening of the grain size away from the top of the flow. Sheet flows recovered at Site U1365 are also noted for their low abundance of chilled margins. Recovery of complete or near-complete flows occurred in Cores 329-U1365E-5R, 6R, 8R, and 12R. Where visible changes in primary groundmass composition and textures took place, the flows were defined as subunits.

The sheet flows range from aphyric to moderately phyric basalt with a groundmass of plagioclase, clinopyroxene ± olivine, and Fe-Ti oxides. Texturally, the lava flows generally vary from subophitic to hyalophitic and glomeroporphyritic. Features observed in individual thin sections include spinifex, spherulitic, ophitic, subophitic, hyalophitic, intersertal, and glomeroporphyritic textures. Grain sizes range from cryptocrystalline to fine grained in the central portions of some flows, with textures that vary from intersertal to intergranular. Vesicle abundance may be highly variable (nonvesciular to highly vesicular) within a single flow (e.g., Subunit 6b) or almost homogeneous (e.g., Subunit 4b). Flow contacts and margins often exhibit increased abundance of vesicles. Almost all vesicles are filled with secondary minerals (see “Basement alteration” for further detail). Phenocrysts include olivine, clinopyroxene, and plagioclase, all of which range in abundance from sparsely phyric to highly phyric. Phenocryst sizes range from 0.2 to 5 mm, with the largest phenocrysts being blastic plagioclase. Flow margins are usually altered and exhibit cryptocrystalline and sometimes glassy textures.

A number of glassy rinds and hyaloclastite fragments, almost entirely composed of altered glass, were recovered in interval 329-U1365E-8R-4, 0–25 cm. The glass exhibits alteration that ranges from slight to complete. Small flakes of vitreous, conchoidal fresh glass are present. In addition, a cryptocrystalline chilled zone that grades to glass is observed within the basaltic groundmass. The glassy zone is interpreted to represent the top of lava flow Unit 13 (massive flow).

Alteration occurs throughout the sheet flows and ranges from slight to complete. Alteration may include groundmass replacement, vesicle fill, veins, halos, and alteration patches (see “Basement alteration”). The number of fractures, breccias, and veins within the sheet flows is highest in the flow margins, whereas the majority of flow interiors remain relatively fresh. The overall low level of alteration and the high abundance of massive flows likely aided overall core recovery. Relatively low recovery occurs between flows and at the flow boundaries where alteration is greatest. Consequently, the majority of the altered rock (and perhaps a large proportion of the breccia) is probably not recovered.

Units 10, 11, and 14–16 (thin flows)

Thin flows occupy 14% of the recovered core (5.8 m). Identification of thin flows was based on the presence of chilled margins, small grain size (microcrystalline to cryptocrystalline), small intervals (tens of centimeters) between each flow, and relatively high alteration extent. Mineralogically, the thin flows are similar in composition to the massive flows, with textures ranging from subophitic to hyalophitic, spinifex, and glomeroporphyritic. Phenocrysts include olivine, clinopyroxene, and plagioclase and range in abundance from sparsely phyric to highly phyric. Vesicle abundance ranges from sparse to vesicular, and most vesicles are filled with various low-temperature secondary minerals (see “Basement alteration”). Textures observed in thin section samples of the thin flows include subophitic, spinifex, hyalophitic, intersertal, and glomeroporphyritic. Thin section observations indicate that most of the thin flow basalt is altered. Alteration ranges from slight to high and is characterized by mesostasis and phenocryst replacement, filling of vesicles, replacement of glassy margins, and vein formation with adjacent alteration halos. Overall alteration is higher in these flows than in the massive flows.

Igneous petrology

As described above, the basaltic rocks recovered from Hole U1365E are divided into massive sheet flows and thin basalt flows. A total of 35 samples were selected for petrographic analyses by thin section (see Site U1365 thin sections in “Core descriptions”).

Massive sheet flow basalt

The mineralogy of the lava flow units at Site U1365 (Units 1–9, 12, 13, and 17) is typical of seafloor basalt. Primary igneous textures and mineralogical differences within the massive flow units broadly fit into three categories: flow top, flow center, and flow base.

The massive flows near the top of the lava flow units are very fine grained (cryptocrystalline to microcrystalline) with a groundmass that is largely composed of plagioclase and clinopyroxene. Accessory Fe-Ti oxides account for 2%–5% of the rock. Phenocryst abundance in the massive sheet flows ranges from 0% to 10%. However, the majority of lava flow tops are aphyric (<0.5% phenocrysts). Where phenocrysts are present, they are composed of blocky to prismatic plagioclase and clinopyroxene crystals that typically range from 0.2 to 1 mm in size. Olivine is rare and only observable as mixed phyllosilicate and iron oxyhydroxide pseudomorphs that can be recognized by a rough six-sided outline; such olivine pseudomorphs are termed “iddingsite.”

Lava flow Units 3, 5, 8, and 9 are sparsely to highly phyric with 0.5%–5% (Units 3 and 5) to 10%–15% (Units 8 and 9) phenocrysts. The most abundant phenocryst phase is plagioclase (90%), followed by clinopyroxene (10%) and olivine (<0.5%). With respect to their groundmass, Units 6, 7, 8, and 9 contain very large (~5 mm) blocky to prismatic plagioclase phenocrysts (Fig. F17). Phenocrysts observed in Sample 329-U1365E-2R-1, 51–53 cm, exhibit complex zoning and twinning patterns. In addition, numerous small inclusions (tecoblasts) are observed within cleavage planes of the plagioclase (Fig. F17). Unaltered inclusions are black, however, most of the inclusions have been altered (they appear brown-green in plane polarized light).

The massive flows at Site U1365 display intersertal textures that range from subophitic to holocrystalline, with interstitial zones frequently holocrystalline. Zones of high vesicle content are also present; they are typically near flow tops, although they also occur at the base of flows (if recovered).

Lava flow Units 1, 2, 4, 6, 7, and 12 share similar mineralogy to Units 3, 5, 8, and 9; however, they contain very few phenocrysts (0% and 0.5%). Phenocrysts in Units 1, 2, 4, 6, 7 and 12 are comprised of prismatic to blocky plagioclase (80%) and subhedral clinopyroxene (20%) that range in size from 0.2 to 1 mm. In Unit 4, phenocryst abundance increases to 0.5% in the central portion of the flow.

A flow contact composed of cryptocrystalline to glassy chill margins with holocrystalline interstitial filling is preserved between two flows is preserved in lava flow Unit 12 (interval 329-U1365E-8R-3, 116–130 cm; Fig. F18). Alteration throughout this interval is slight with only minor saponite/celadonite in the groundmass and subhorizontal late-stage carbonate veins. Consequently, the full contact is preserved. The lack of alteration and complete preservation of the contact implies that the subsequent flow was rapid, reducing the period of exposure to open to seafloor weathering.

Thin basaltic flows

The thin basalt flows (lava flow Units 10, 11, and 14–16) are similar in composition to the massive flows, with plagioclase (58%–70%), clinopyroxene (26%–35%), Fe-Ti oxides (2%–5%), and rare olivine (~<1%) making up the groundmass. The plagioclase typically forms bladed crystals that are intergrown with anhedral to subhedral clinopyroxene and Fe-Ti oxides (titanomagnetite). The groundmass ranges from cryptocrystalline to microcrystalline, and the most common textures are subophitic to hyalophitic. Grain size within the thin flows varies at a localized scale (millimeter to centimeter) where grain size changes from microcrystalline to cryptocrystalline to glassy at chill margins and flow contacts. However, within each flow, no changes were observed. Textures close to the flow margins are typically variolitic to hyalophitic. Phenocryst abundance in Unit 10 and in the top portion of Unit 11 is 10% (highly phyric); these phenocrysts are entirely composed of clustered prismatic to blastic plagioclase that range in size from 0.2 to 6 mm. Units 14–16 are aphyric. Vesicle abundance ranges from none to ~6%. Units 11, 14, and 16 contain 1%, 1.5%, and 6% vesicles, respectively. The majority of the vesicles are concentrated at the top of the flows, near the chill margins. Units 10 and 15 are vesicle free.

Phenocryst phases

Plagioclase

Plagioclase phenocrysts are present throughout the basement at Site U1365. Plagioclase phenocrysts make up ~2% of the massive basalt flows and ~1% of the thin flows, making plagioclase the most abundant phenocryst phase. Plagioclase phenocrysts are euhedral to subhedral in shape. Although they range from 0.1 to 7 mm in length, most are between 0.5 and 2 mm. Rarely, plagioclase forms skeletal or quench plagioclase crystals in lower pillow lava units. Zoning of plagioclase is relatively common and is more prominent in the larger phenocrysts (Fig. F17). Blebs and microlites of glass and clinopyroxene occasionally form inclusions that run parallel to twinning planes in some phenocrysts. Although plagioclase phenocrysts are typically fresh, replacement by secondary minerals can vary from 0% to 50%. Replacement minerals that include clays, saponite, and iron oxyhydroxides occur along cracks, cleavage planes, or crystal edges (Fig. F19).

Pyroxene

Clinopyroxene phenocrysts are present throughout the Site U1365 basalt and make up ~0.5% of the recovered core. These phenocrysts range from 0.2 to 2 mm in length and are typically anhedral, ranging from round to angular, with simple basal twinning present throughout. Clinopyroxene is typically intergrown with plagioclase in glomeroporphyritic clots or as subophitic crystals around plagioclase. Alteration of clinopyroxene varies from 0% to 80% and manifests as replacement by secondary clays, saponite, iron oxyhydroxides, and oxides along cracks, cleavage planes, or crystal edges.

Olivine

Fresh olivine is extremely rare, and olivine pseudomorphs make up <0.1% of the recovered material. Olivine phenocrysts are, on average, 0.2 mm wide and are almost always completely replaced by secondary minerals. Secondary mineral replacement usually consists of iddingsite, but celadonite, and opaques (sphene and Fe-Ti oxides) are also observed. Their identification therefore relied on their crystal morphology (subhedral to euhedral) and their textural relationships with surrounding minerals. Larger olivine phenocrysts in flow interiors are typically skeletal in structure and consist almost entirely of replacement minerals.

Groundmass

The basaltic groundmass at Site U1365 varies from hypocrystalline to holocrystalline and is composed primarily of plagioclase and clinopyroxene, with minor accessory Fe-Ti oxides. Olivine is rare. Plagioclase is the most abundant groundmass crystalline phase, comprising between 58% and 70% of the groundmass. Plagioclase occurs as microlaths, microlites, and acicular crystals in a spinifex texture or microcrysts in chill margins. Clinopyroxene is the next most abundant primary igneous phase after plagioclase, and it comprises approximately 35% of massive sheet flows and ~30% of thin basalt flows. Clinopyroxene occurs as interstitial growths between plagioclase, microlaths, microlites, and aggregates of fibrous or plumose crystals. Anhedral to subhedral microcrysts of olivine pseudomorphs are present in low abundance in the massive lavas and thin basalt flows. The lack of fresh olivine and the difficulty in identifying olivine pseudomorphs (based on relict crystal structure) hamper efforts to estimate the abundance of olivine; however, our observations imply that olivine abundance ranges from 0% to 1%. Mesostasis at Site U1365 ranges from 0.5% to 4% and is present throughout the recovered basement. Mesostasis textures include hyalophitic, intersertal, spherulitic, and variolitic. Mesostasis within the thin basalt flows and near chill margins is dominated by spherulitic and variolitic textures

Mesostasis is typically subject to patchy alteration, in which it is preferentially altered relative to the plagioclase and clinopyroxene groundmass. Almost all patchy alteration observed at Site U1365 is the result of altered mesostasis. Replacement minerals in the groundmass include clay (saponite and celadonite), iron oxyhydroxides, and, rarely, carbonate. Primary magmatic opaques (<1%–4%) are present in all units. These form small (<0.2 mm), granular, partially replaced, subhedral crystals of titanomagnetite. Vesicles are typically present near or at the tops and bottoms of the massive lava flows; however, they may be present throughout any one flow, particularly in the thin flows. Most vesicles (>60%) are completely filled by mono- to polymineralic secondary assemblages, including saponite, celadonite, iron oxyhydroxides, pyrite, and mixed clay. Further details regarding secondary mineral vesicle fill are described in “Basement alteration.”

Hard rock geochemistry

Sixteen representative samples of the basaltic basement were analyzed for major and trace elements using a Teledyne-Leeman (Prodigy) inductively coupled plasma–atomic emission spectrometer (ICP-AES). The representative samples include relatively unaltered basalt groundmass (gray to green) and variably altered halos (red to brown). The least altered samples were chosen based on the lowest abundance of secondary mineral phases present in thin section and the least number of veins, halos, and filled vesicles. Altered samples were chosen to ensure that each alteration phase is represented. Details of the methods for preparation and analyses are detailed in “Lithostratigraphy, igneous petrology, alteration, and structural geology” in the “Methods” chapter (Expedition 329 Scientists, 2011). International standard BCR-2 was analyzed 24 times over 3 runs. The analytical precision and accuracy is reported in Table T3 in the “Methods” chapter (Expedition 329 Scientists, 2011).

Results

Major and trace element data and loss on ignition (LOI) for the selected samples are shown in Table T2. For all basaltic samples, major element oxide contents include

  • SiO2 = 49.3–52.8 wt%,

  • Al2O3 = 13.6–20.2 wt%,

  • Fe2O3 = 7.2–10.5 wt%,

  • MgO = 6.3–8.1 wt%,

  • Na2O = 2.3–3.0 wt%,

  • TiO2 = 0.93–2.1 wt%, and

  • K2O = 0.04–1.33 wt%.

Trace element contents and averages include

  • Sr = 112–166 ppm (average = 138 ppm),

  • V = 195–424 ppm (average = 316 ppm), and

  • Zr = 39–118 ppm (average = 86 ppm).

Trends in least altered basalt

Eight least altered samples that ranged in color from gray to gray-green were selected, based on their low abundance of secondary minerals, for primary whole-rock chemical analyses. Total alkaline (K2O + Na2O) ranges from 2.5 to 3.7 wt%, and SiO2 ranges from 50 to 52 wt%. Al2O3 ranges from 14 to 20 wt%, and CaO ranges from 11.0 to 13.8 wt%. Samples 329-U1365E-3R-4, 67–68 cm; 7R-2, 5–9 cm; and 8R-2, 54–58 cm, all exhibit high Al2O3 (17.6–20.2 wt%) and CaO content (12.7–13.8 wt%). These samples contain abundant plagioclase, which may explain the high Al and Ca content.

Downhole variation in major and trace element concentrations in the basalt are shown in Figure F20. MgO and Sr contents decrease with increasing depth, whereas MnO2, Fe2O3, Ge, Zn, Ba, and V all increase with increasing depths. K2O/TiO2 ratio at Site U1365 ranges from 0.02 to 0.42, indicating either a range from depleted to enriched basaltic compositions or relative distribution of potassium-rich secondary minerals (e.g., saponite and celadonite).

Relationships of MgO with some major elements and incompatible and compatible trace elements are shown in Figure F21. Na2O, Fe2O3, and TiO2 increase with decreasing MgO.

The extent to which downhole variation in chemical composition of these least altered samples is due to magma evolution or basalt alteration will be addressed by postexpedition research.

Basement alteration

All basement rocks at Site U1365 have been subjected to alteration by interaction with seawater. Alteration varies from slight to complete. However, the majority of recovered basement material at Site U1365 is only slightly altered. Basement alteration at Site U1365 is characterized by

  • Replacement of phenocrysts by secondary mineral assemblages,

  • Replacement of mesostasis in the groundmass by secondary minerals,

  • Filling of veins and the formation of halos by emplacement of secondary minerals, and

  • Lining and filling of vesicles.

Visible alteration in macroscopic view or thin section constitutes between 2% and ~95% of individual samples, with most alteration concentrated around veins and vesicles. The most intense alteration is present in and/or near flow margins and breccias, where alteration can be nearly complete. Alteration products include saponite, celadonite, iron oxyhydroxides, carbonate, and accessory zeolite (laumontite by XRD). Other secondary minerals include sulfides (chalcopyrite and pyrite) and quartz.

Identification of secondary minerals during Expedition 329 was carried out by macroscopic observation, thin section observation, and XRD analyses (Table T3). Clay minerals are predominantly in the saponite group or celadonite group; these were distinguished by color variations.

Saponite is pervasive throughout the recovered core. In macroscopic observation, colors range from black, dark green, greenish brown, or pale blue. In thin section, it is characterized by a pale brown color and may be mottled or fibrous in form. Replacement of the groundmass is usually even and slight, replacing olivine, mesostasis and some of the groundmass. In areas of moderate alteration, saponite replaces mesostasis and a varying proportion of groundmass crystals. Highly to completely altered basalt exhibits continuous mottled replacement, destroying most or all original textures (Fig. F22). Saponite also frequently fills vesicles, forms monomineralic or polymineralic veins, and is a component in breccia matrixes.

Celadonite is present throughout the recovered basement at lower abundance than saponite. In hand specimen and thin section, celadonite is distinctively bright green-blue and it typically fills veins and vesicles and replaces primary interstitial zones in basaltic groundmass.

Iron oxyhydroxides are the next most abundant alteration mineral after celadonite; they are present throughout the recovered core. Iron oxyhydroxides can occur alone or, more commonly, intermixed with saponite, imparting red to brown staining to the saponite. Iron oxyhydroxides are identifiable by a bright red-orange color, and they often stain other secondary mineral phases. In addition, iron oxyhydroxide typically replaces phenocrysts as iddingsite to form hyalophitic texture. Iron oxyhydroxides may also fill or partially fill veins, and they commonly form iron oxyhydroxide–dominated halos.

Less dominant secondary minerals at Site U1365 identified by XRD include calcite, zeolite (laumontite and phillipsite), and clay (montmorillonite and sepiolite). Although calcite is rare in the basaltic groundmass, it is the most dominant vein-filling mineral and forms a major constituent in breccia matrixes, alteration patches, and vesicles.

Pyrite identified at Site U1365 is gold-colored in macroscopic observation and bright yellow in reflected-light microscopic observation. Pyrite occupies alteration patches and veins and occasionally replaces interstitial zones within halos or along halo boundaries. Its crystal structure ranges from blocky to amorphous, and poor cleavage is sometimes observed. Figure F23 highlights examples of pyrite. The vast majority of pyrite occurs in the lower portion of the hole (i.e., Cores 329-U1365E-11R and 12R).

Alteration features in the basement at Site U1365 are described below in order of alteration intensity. At Site U1365, breccias exhibit the greatest degree of alteration, albeit at very concentrated zones. The most pervasive forms of alteration at Site U1365 are halos and veins. These features occur throughout the recovered core and represent the evolution of low-temperature secondary mineral emplacement within the basalt. Vesicle fill is perhaps the least pervasive form of alteration. However, it provides a clear indication of the relative timing of secondary mineral emplacement and, as such, is described separately.

Breccias

Breccias at Site U1365 can be divided into three types:

  • Hyaloclastite (or magmatic) breccia,

  • Basaltic breccia, and

  • Mixed breccia.

Breccia makes up <0.5% of the recovered core. However, because of the preferential recovery of rheologically stronger units (sheet flows and massive flows), it is likely that the recovered breccia percentage underestimates the true proportion of brecciated basement at Site U1365.

Hyaloclastites were recovered in Sections 329-U1365E-8R-3 and 8R-1. These consist of fresh and altered glassy clasts in a phyllosilicate and/or carbonate matrix. The clasts are angular to subrounded and range in size from 0.4 to 20 mm. Alteration in this breccia ranges from moderate to complete, with pervasive alteration throughout the clasts and multiple phases of veins. A number of veins protrude into the clasts from the matrix and are therefore contemporaneous with the formation of the breccia. The clasts are variably altered to saponite (yellow-brown with low birefringence), whereas the centers of the clasts represent either mixtures of hydrated glass or less intense saponite alteration. Alteration zones in individual glass clasts indicate that the clasts were originally angular but alteration has replaced the corners and edges, leaving subangular to rounded blunt shards (Fig. F24A, F24C). Larger clasts (>15 mm) are less altered and more angular in shape, implying that these clasts brecciated at a late stage. A small number of plagioclase phenocrysts in the glass range from fresh to partially altered. Within the matrix, a number of chilled curvilinear basaltic clasts are present. In addition, 1–5 mm sized clasts composed entirely of clay are present; these are inferred to be entirely replaced glass.

Basaltic breccia was recovered in Section 329-U1365E-5R-4. This breccia is composed of submillimeter to several centimeter subangular to angular basaltic clasts with a matrix of carbonates and saponite (Fig. F24A). Multiple infills of secondary minerals and the formation of numerous veins, further widening the gaps between the basalt fragments, suggest that these breccias formed in situ as the end-member of a vein net. The clasts exhibit variable slight to high alteration, which manifests itself in the form of multiple halos, vesicle fills, and veins intruding into the clasts. The clasts comprise cryptocrystalline to fine-grained basalt that ranges from hypocrystalline to holocrystalline. The clasts are mineralogically typical for a basaltic assemblage, with plagioclase, clinopyroxene, olivine, and minor accessory Fe-Ti oxides (in order of abundance) making up the groundmass. Plagioclase occurs as microlaths, microlites, plumose acicular crystals, and quench crystals and is marginally the most abundant crystalline phase in the groundmass. Clinopyroxene within the basaltic clasts occurs as interstitial growths between plagioclase crystals. Clinopyroxene may occur as microlaths, microlites, and aggregates of fibrous or plumose crystals. Phenocrysts make up <0.5% of the clasts and are composed of plagioclase, clinopyroxene, and olivine pseudomorphs. Partial replacement of plagioclase and clinopyroxene phenocrysts takes place within intraclast alteration halos and toward the edges of the clasts. In Core 329-U1365E-8R, the distribution and abundance of each phenocryst phase within the clasts is similar to their distribution and abundance in igneous Unit 12 (see “Igneous petrology” for details).

Mixed breccias at Site U1365 are composed of glass and basaltic clasts in a matrix of silicates, basaltic fragments, glassy shards, and calcite. The one example of this breccia type recovered from Site U1365 is located in interval 329-U1365E-5R-4, 59–63 cm (Fig. F24B). The breccia is composed of basaltic clasts with a matrix formed from glass that intruded into the rock. The glass is in turn altered and fractured, with a secondary matrix of carbonate, celadonite, saponite, and minor iron oxyhydroxides. The primary mineralogy and igneous texture of basaltic clasts in the mixed breccia are identical to mineralogy and texture in the adjacent igneous unit (Unit 5) (described in “Igneous petrology”). The basaltic clasts are angular to subangular and poorly sorted with alteration that ranges from slight to high. Zoning of alteration within the clasts is characterized by alteration halos that form around the clast edges. The clasts are pervasively altered to saponite and iron oxyhydroxides. Zoning of alteration toward the chilled margins in the clasts implies that initial alteration took place prior to brecciation.

Vein and halo-related alteration

Dark gray/brown saponitic and celadonitic background alteration occurs throughout recovered basement at Site U1365. However, vein-related alteration is also present throughout the site as localized, variably colored alteration halos along veins that are composed of a variety of secondary minerals. In many cases, the alteration halos are preserved where the vein was not recovered. Veins may be monomineralic or polymineralic and may contain any combination of the following secondary minerals: saponite, celadonite, iron oxyhydroxides, calcite, and accessory zeolite, secondary sulfides, and silicates (quartz and chalcedony). Alteration halos along the vein margins at Site U1365 include dark green/black halos, green-brown halos, red halos, and mixed halos. The most abundant halo type is red-brown to orange-brown, followed by dark gray halos. Red halos and complex (mixed) halos are the least common and primarily occur near flow margins. Halos and veins are discussed in detail below.

Halos

Dark green/black halos are present throughout Site U1365. However, they are most concentrated at the top and base of each igneous unit and in Core 329-U1365E-12R. The term dark green/black halos refers to all halos that vary from very dark gray to dark green (Fig. F25A, F25B) and may be incorporated into mixed halos (e.g., interval 329-U1365E-8R-1, 40–48 cm; Fig. F25D). Dark green/black halos range in width from 1 to 25 mm, but most are commonly 1–10 mm wide. Secondary mineral abundance in the dark green/black halos is usually similar or slightly greater than that of gray background. However, the mineralogy of dark green/black halos is characterized by celadonite replacing olivine and mesostasis and filling vesicles. Celadonite is identified by its green color in thin section or its blue-green color and brittle texture in hand specimen. Celadonite typically replaces between 2% and 5% of the rock within the halo. Our estimate of celadonite is based on visual observation by hand specimen, thin section identification, and shipboard XRD analyses. Since the majority of celadonite identification was based on visual observation, mineral phases identified shipboard as celadonite may include celadonite, nontronite, and mixed-layered celadonite-nontronite. Detailed XRD analyses will be required to refine identification further. Within the dark green/black halos, saponite is observed replacing olivine, mesostasis, and vesicles, as well as overprinting celadonite. Vein and vesicle filling sequences (discussed later) indicate that the saponite phase arrives after celadonite. Iron oxyhydroxides may also be present in small amounts.

Dark green/brown halos occur throughout the recovered core at Site U1365. However the greatest intensity of these halos occurs at flow boundaries and in Core 329-U1365E-12R. Dark green/brown halos range in width from 2 to 30 mm. However, the majority of these halos range from 5 to 13 mm in width. These halos are typically associated with saponite veins, but they also flank polymineralic veins and veins of celadonite, iron oxyhydroxides, and/or carbonate (Fig. F25A, F25B). Thin section observation indicates that the dominant secondary mineral is saponite, which is green to brown in plane-polarized light and dark green/brown in hand specimen. Saponite fills vesicles and replaces olivine phenocrysts and interstitial material and it comprises between 3% and 80% of the total rock within the halo. Most halos, however, exhibit only slight to moderate (3%–20%) replacement by saponite. Frequently present within these halos are iron oxyhydroxides (0.5% to 2% of the rock within the halo) that stain the saponite to orange-brown.

Red halos represent a range of colors, including brown, orange, and red, which are distinguished by a high proportion of iron oxyhydroxides. Other mineral phases that are sometimes present include saponite, celadonite, and carbonate. Red halos occur in greatest concentrations at the top and base of each flow unit. In Cores 329-U1365E-2R through 10R, red halos are absent in the centers of the units. Iron oxyhydroxides make up between 3% and 10% of the rock in the halo and fill vesicles and replace olivine and interstitial areas. An example of an iron oxyhydroxide halo is shown in Figure F25C. Careful observation reveals that iron oxyhydroxide occurs as narrow strands that propagate between individual grain boundaries and thus stain the background rock to form the halo. Red halos range in thickness from 1 to 15 mm. Within the halos, concentrations of iron oxyhydroxides commonly form very dark red/brown bands with halo margins typically irregular. Iron oxyhydroxide may form concentrated zones within the halo. Red halos are typically associated with veins of iron oxyhydroxide or iron oxyhydroxide and celadonite, but they can also surround saponite and carbonate veins.

Mixed halos occur almost exclusively near flow contacts and unit boundaries and are not commonly present within the flow centers. Mixed halos are the result of multiple overprinting stages from dark green/black halos, red halos, and green-brown halos. Typically, only two zones occur. However, there are several samples that exhibit the complete sequence of halos. Interval 329-U1365E-8R-1, 40–60 cm, exhibits a very complex sequence in which several vein-filling generations and subsequent halo emplacements took place (Fig. F26). Mixed halos are between 10 and 40 mm wide, whereas individual halos within each set of halos range in width from 1 to 20 mm. The mineralogy of each individual halo within each mixed halo essentially falls into any one of the dark green/black, red, or green-brown halo categories. However, because of overprinting, the innermost halo usually contains mineralogy that relates to two or more alteration phases; therefore, the coloration will be mixed. As with the simple “single alteration phase” halos, the intensity of coloration reflects the level of alteration. In mixed halos, the dark green/black celadonitic halos are partially to completely overprinted by saponite and iron oxyhydroxides. In most mixed halos, only discrete patches of celadonite remain. In a number of mixed halos, earlier sequences can be overprinted by later alteration halos that extend well beyond the boundary of the previous halo. Multilayered halos may also exhibit patchy and indistinct alteration fronts (e.g., interval 329-U1365E-8R-1, 40–60 cm; Fig. F27). In this interval, halos extend laterally in a series of frond-like structures from a multimineralic vein. This unusual pattern may be the result of relatively weak flow planes that allow greater propagation of secondary mineral emplacement.

Veins

A total of 593 veins were identified in the basement core recovered from Hole U1365E, with an average density of 14 veins/m of recovered core (Table T4). Vein fill makes up 1.23% by volume of recovered core. Vein thickness varies from <0.1 to 10 mm, although most veins fall are in the 0.1–1 mm range. Veins observed in basement at Site U1365 exhibit planar, straight, curved, branching, anastomosing, kinked, sinusoidal, irregular, and crosscutting morphologies. Secondary minerals that fill veins include saponite, celadonite, iron oxyhydroxides, carbonate, and accessory phases (other unidentified clays, quartz, chalcedony, zeolite, and secondary sulfides). Veins may be monomineralic or polymineralic, with any combination of the major secondary minerals. Veins may be flanked by alteration halos, or, rarely, they may simply penetrate the groundmass with no alteration halo. Crosscutting relationships and vein-filling orders, relative to each vein mineral, are described below.

Saponite is present in nearly all the veins and makes up 0.04% by volume of the recovered core. Saponite-bearing veins range from <0.1 to 3 mm thick, and their average thickness is ~0.2 mm. Saponite is observed to occur with every other secondary mineral; however, it is most commonly associated with iron oxyhydroxides and carbonate. Saponite typically crosscuts celadonite and iron oxyhydroxide and is itself crosscut by carbonate and, rarely, by zeolite and silicates. Suitable samples for XRD analyses to identify saponite were rare because of either low sample volume or the presence of multiple mineral phases that could cause contamination. The one saponite vein that was analyzed by XRD indicates a trioctahedral smectite structure.

Iron oxyhydroxide is also present in many veins at Site U1365. Iron oxyhydroxide comprises 4.2% of all veins and makes up 0.2% of the recovered core (Table T4). Iron oxyhydroxide veins range from <0.1 to 4 mm thick (0.3 mm thick on average). Although a number of veins are exclusively iron oxyhydroxide (e.g., in Figs. F27A, F27E, F22B), most iron oxyhydroxide is present with saponite, celadonite, and calcite. Iron oxyhydroxide is typically overprinted or crosscut by saponite and calcite; however, it is often overprinted or intergrown with celadonite. Iron oxyhydroxide is present throughout Hole U1365E. However, vein abundances are greater in the uppermost two-thirds of the hole. These veins, like the iron oxyhydroxide–rich halos, are uncommon within flow centers.

Celadonite-bearing veins comprise 16% of the total number of veins at Site U1365 and form 0.2% of the recovered core by volume (Table T4). Vein thicknesses vary from <0.1 mm to occasionally spectacular 10 mm veins. Most celadonite-bearing veins are between <0.1 and 0.3 mm thick, and pure (100%) celadonite veins tend to be narrow (<0.1–0.3 mm thick). Most celadonite veins are either intergrown with or overprinted by iron oxyhydroxides and are largely overprinted by saponite and carbonate. In many veins, only discrete patches of celadonite remain. Celadonite was identified in thin section by its green color and in hand specimen by its blue-green color and brittle texture. XRD analyses of celadonite indicate intergrowths of saponite.

Carbonate is the most commonly occurring mineral phase in veins; it occurs in 60% of the total number of veins and it makes up 0.74% of the total volume of recovered core. Carbonate is present in its own veins, crosscutting celadonite, iron oxyhydroxide, and saponite, or, more frequently, as a late-stage infill in polymineralic veins (Fig. F27A–F27D). The proportion of carbonate in a given polymineralic vein ranges from a trace (<0.5%) to almost 100%. In almost all situations where overprinting/replacement relationships can be discerned, carbonate replaces all other major secondary phases. Veins bearing carbonate can be anywhere from <0.1 to 20 mm thick. Most pure (100%) carbonate veins are not flanked by halos. A number of vertical to subvertical veins with only carbonate infilling occur with no halos. These veins usually have no halos flanking them and appear to crosscut all other subhorizontal veins, including carbonate.

A number of ultrafine phyllosilicates within some vein material are present in some veins. However, they are too fine to identify by thin section and too small to be sampled by XRD. These samples may represent mixed interlayered clays. In addition, rare 0.1 mm thick sulfide veins and patches are present toward the base of Hole U1365E (Fig. F23).

Vesicles

All units in Hole U1365E contain vesicles, the abundance of which varies from <0.1% to 20%. Most vesicles are partially to totally filled with one or more secondary minerals. Thin section observations indicate that most vesicles are 100% filled. Secondary minerals in vesicles include saponite, celadonite, iron oxyhydroxides, and calcite, in order of occurrence. Vesicle fill is highly variable in each unit. Sometimes, individual vesicles may contain one to three different secondary minerals. On both flow unit and piece scales, the variability of vesicle-filling minerals is high, with the typical assemblage of each unit containing seven or more different secondary minerals. Vesicles within alteration halos are usually filled with the dominant phase of that halo (e.g., iron oxyhydroxides in a red halo). However, earlier fill (lining at the edge of a vesicle) may also be present (Fig. F28). In order of filling, common mineralogical relationships within vesicles observed at Site U1365 are

  • Celadonite, saponite, and calcite;

  • Saponite and calcite;

  • Iron oxyhydroxide and calcite;

  • Iron oxyhydroxide and saponite;

  • Celadonite, saponite, and calcite; and

  • Calcite (generation 1) and calcite (generation 2).

The high variability of vesicle fill history and vesicle fill distribution indicates that continuous, localized fluid evolution and secondary mineral emplacement has taken place. Veins and halos record widespread major alteration phases that are pervasive throughout the cored basement of Site U1365.

Black calcite

Three pieces of black opaque calcite were recovered at Site U1365 (Fig. F29). The pieces were identified as calcite by effervescence on application of 5% HCl solution and by rhombic crystalline structure within small vugs. The black calcite occurs at intervals 329-U1365E-5R-1, 0–4 cm, and 6R-1, 14–26 cm (in Units 4 and 7, respectively). Thin section observation indicates that the crystalline mass is actually composed of colorless calcite that exhibits perfect basal cleavage, moderate relief, and high birefringence (fourth-order pale pink and green coloration). The black color, however, derives from numerous inclusions, opaque minerals, and other tiny <0.01 mm minerals. Noncalcitic material makes up ~30% of the total volume of the black calcite sample. Inclusions are ~10% of the overall calcite and are typically filled with irregular isotropic opaque minerals. Rare clear inclusions are present that may contain fluid or gas. Approximately 3% of the calcite is composed of a highly reflective tiny (<0.02 mm) opaque mineral that shows similar characteristics to marcasite (white to slightly yellow lathlike crystals). A number of other opaque minerals yet to be identified are present. Individual crystals (<0.02 mm in length) of zeolite comprise 2% of the calcite. These crystals are colorless in plane-polarized light and are low relief and twinned. The position of these calcite pieces near lithologic unit boundaries and their association with generally high levels of alteration suggest that they formed as part of large interflow alteration zones. ICP-AES measurements of the black calcite (Table T2) show that it has high silica (SiO2 = 22 wt%), Fe (Fe2O3 = 9 wt%), and Ba (662 ppm) contents. The presence of silica suggests that much of the noncalcitic material may be remnant basaltic groundmass. Iron content (as measured by ICP-AES) may indicate that opaque ferrous minerals such as titanomagnetite might make up the other secondary minerals present. The presence of Ba may indicate substitution of Ca with portions of the calcite crystal lattice to form witherite (BaCO3).

Biogenic alteration features

A number of tubelike, micrometer-scale weathering features of potential microbial origin are observed in Sample 329-U1365E-8R-4, 3–6 cm (Fig. F30). Tube morphologies include irregular, branching, spiraling, and segmented and they range in size from 0.5 to 5 µm in diameter. The tube diameter remains constant throughout the majority of the tubes, including branched sections. Tubes are arranged either in discrete clusters or in masses adjacent to or near fractures and iron oxyhydroxide within altered glass. Rare spherical, dark/opaque inclusions (0.2–0.5 µm) are observed within several tube structures. The morphology, size, and location of tubes observed in altered glass at Site U1365 appear similar to biogenic alteration features that have been previously observed in marine basaltic glass (Fisk et al., 1998).

Compositional comparison of alteration features to least altered material

A small suite of seven samples was selected for shipboard study of compositional alteration at Site U1365. The ICP-AES results are presented in Table T2. The altered samples were selected based on visual observation of secondary minerals within the groundmass, either as alteration halos or as alteration present within the groundmass. Four of the alteration halo samples were selected to have direct contact with a measured “least altered” background (Samples 329-U1365E-2R-2, 30–34 cm; 3R-4, 67–78 cm; 8R-2, 59–61 cm; and 11R-3, 123–126 cm). Two other samples, 4R-1, 25–27 cm, and 2R-2, 30–34 cm, were paired with Samples 4R-1, 85–87 cm, and 2R-1, 51–53 cm, respectively. For these two samples, proximity and visual similarity with fresh contacts were inferred to be suitable substitutes for actual contacts; their actual fresh contacts were not sampled because of sample volume limitations and risk of contamination induced by small sample size (e.g., irregular halo margins, closeness to core boundary, and narrowness of fresh portion). The major and trace element concentrations of these sample pairings are reported in Table T2.

Ranges and averages of some key elements for altered samples include

  • Fe2O3(T) = 7.12–14.0 wt% (average = 10.53 wt%),

  • MgO = 6.29–8.05 wt% (average = 7.13 wt%),

  • CaO = 11.16–13.50 wt% (average = 12.57 wt%),

  • K2O = 0.10–1.33 wt% (average = 0.62 wt%),

  • TiO2 = 0.93–1.98 wt% (average = 1.48 wt%), and

  • Sr = 112–162 wt% (average = 142 wt%).

Overall differences between the ranges and averages of the altered and the least altered basalt are relatively small. On average, an overall increase from least altered basalt to altered basalt occurs with Al2O3, K2O, Ba, Sr, and LOI. Overall average decreases include Fe2O3(T), MgO, TiO2, P2O5, Co, V, and Zr. These average differences between least altered and altered basalt may reflect variable replacement of groundmass by secondary minerals and scavenging of metals (including Fe) to form secondary minerals within veins (e.g., corrosion and/or replacement of magnetite to supply iron oxyhydroxides in veins). Minimal differences in Ca and Mg between most altered basalt and least altered basalt suggest that Ca/Mg exchange between seawater and wall rock has been either modest or pervasive in the recovered basalt.

For basic assessment of elemental mobility within the whole rock, ratios of altered versus unaltered rock for sample pairings are shown in Figure F31A. Sample pairings for halos that were not directly associated but share similar relationship are shown in Figure F31B. All samples from Site U1365, except Sample 329-U1365E-8R-4, 59–61 cm, versus 8R-4, 61–64 cm, show slight to large increases in LOI. Elemental changes observed in both the actual and chosen sample pairs include increased K2O. Samples 2R-1, 30–34 cm, versus 2R-1, 51–53 cm; 3R-4, 67–68 cm, versus 3R-4, 68–69 cm; and 8R-4, 59–61 cm, versus 8R-4, 61–64 cm, exhibit increases in Fe2O3(T) and MnO2 and slight decreases in MgO, P2O5, Co, Cu, Ni, and Sr. Decreases in Fe2O3(T), MnO2, Co, and Cr are observed in Samples 4R-1, 25–27 cm, versus 4R-1, 85–87 cm, and 11R-3, 123–126 cm.

The increases in Fe2O3(T), MnO2, K2O, and LOI may reflect the incorporation of secondary minerals (saponite, celadonite, and iron oxyhydroxides) that contain Fe, Mn, K, and LOI (as water-bound in interlayer sites) into the groundmass. The reduction in Fe2O3(T) and MnO2 in Samples 329-U1365E-4R-1, 25–27 cm, versus 4R-1, 85–87 cm, and 11R-3, 123–126 cm, may have resulted from (1) Fe and Mn scavenging from primary oxides within the groundmass, followed by subsequent precipitation within veins and fractures, or (2) variation in primary composition on a scale larger than that of the sample. In either case, this trend invites postexpedition study.

Overall, the relative changes observed in comparison of altered samples and least altered samples are consistent with (1) incorporation of the secondary minerals saponite, celadonite, and iron oxyhydroxides and (2) partial chemical exchange with the basement. The high variability of trends associated with alteration downhole suggests that alteration varies on a localized scale. It is possible that all rock at Site U1365, including the least altered rock, has undergone some degree of alteration. Detailed postexpedition work will be required to fully compositionally characterize rock alteration at Site U1365.

Alteration summary

Low-temperature hydrothermal alteration at Site U1365 is similar to the alteration in the uppermost portion of the oceanic basement at other areas where in situ ocean crust has been recovered (e.g., ODP Holes 504B and 1256D and the nearest sites to Site U1365 [DSDP Sites 595/596]) (Shipboard Scientific Party, 1987; Laverne et al., 1996; Teagle et al., 1996, 2006).

Alteration extent was recorded by visual observation from core descriptions and by natural gamma ray (NGR) logging (using spectral NGR-based potassium concentration as a proxy for alteration extent). The visual record and the NGR potassium show strong correspondence. The direct relationship between visual observations of alteration and NGR-based potassium content indicates that NGR can provide a more accurate and quantitative approach to estimating alteration extent at basement sites than visual interpretation alone.

The relationship between igneous unit boundaries and extent of alteration (Fig. F32) indicates that alteration is strongly controlled by the structure of the crust. At Site U1365, ingress of seawater, secondary mineral precipitation, and chemical wall rock interaction is restricted to interflow regions.

Alteration at Site U1365 can be divided into two components: (1) open circulation of seawater causing oxidative alteration and (2) restricted fluid circulation giving rise to oxygen-starved alteration (Laverne et al., 1996; Teagle et al., 1996, 2006). Secondary iron oxyhydroxide and celadonite are typical of oxidative alteration, whereas the presence of saponite and secondary sulfides suggests oxygen-poor alteration. It is not yet clear how these stages are distributed within the basement. Observations of individual veins and halos indicate that, at least on a local level, fluid flow becomes restricted as voids are filled by secondary phases. Localized complex alteration at Site U1365 suggests that some areas have undergone several stages of vein reopening and new halo emplacement (Fig. F26). However, it is still unclear whether renewed oxidative alteration has taken place within these zones.

Restrictive fluid flow leading to these alteration characteristics is most evident toward the base of Hole U1365E because saponite and secondary sulfides within veins, halos, and alteration patches are more prevalent here. The lack of iron oxyhydroxide and relatively low abundance of celadonite in veins and halos in the central portions of the flows suggest oxidative alteration was very limited. In these less permeable zones, fluid flow is likely to be restricted and the zones likely rapidly closed to open oxidizing circulation very shortly after celadonite precipitated.

Our shipboard studies of hand specimens and thin sections provide no evidence that late-stage fills are oxidative (i.e., we observed no late-stage alteration halos, celadonite, or iron oxyhydroxide). However, the presence of dissolved oxygen in the lowermost sediment at below-deepwater concentrations indicates that oxidation continues to take place, albeit perhaps at a very low rate (see “Biogeochemistry”). Late-stage alteration at Site U1365 appears to be dominated by multiple episodes of carbonate precipitation and vein infill. These episodes are evident in a number of veins in which crystal growths exhibit a break in their structure from reopening that is later filled with additional rows of crystals flanking the interior of the vein (e.g., in Sample 329-U1365E-4R-1, 7 cm; Fig. F33). The presence of dissolved Mg in the lowermost sediment at below-deepwater concentrations and dissolved Ca at above-deepwater concentrations indicates that basalt-water interaction in the form of Mg exchange for Ca persists today (see “Biogeochemistry”). This evidence of continued exchange suggests that secondary carbonate precipitation is ongoing. Whether alteration was continuous or occurred intermittently throughout the history of Site U1365 basement remains speculative. However, the presence of late-stage vertical carbonate veins suggest tectonic processes vertically fractured basement, allowing seawater-derived carbonate to precipitate.

Structural geology

Our expedition goals for Site U1365 focused on describing the basalt in terms of its habitability (alteration) and providing a general characterization of the primary features of the host rock. However, we were also able to provide some basic description of the main structural features at Site U1365.

Basalt recovered from Site U1365 has been subject to a small variety of synmagmatic and postmagmatic structural changes, with structural features that include flow laminations, planar flow margins, microfaults, jointing, interflow breccia, veins, and extensional mineral growths. Additional structural features include joints induced by coring and core-handling processes. Structural features were described and entered into the Laboratory Information Management System database through the DESClogik software (see “Lithostratigraphy, igneous petrology, alteration, and structural geology” in the “Methods” chapter [Expedition 329 Scientists, 2011]). In addition, breccia units were described in terms of textural features and composition.

Only geological features were recorded in the standard graphic report, and only planar features were entered. This practice restricted the number of records to only a few measurements.

Veins

The most pervasive and numerous structural features observed within basement recovered at Site U1365A are veins. Structurally, veins are extensional fractures that have been filled with secondary minerals (see “Basement alteration”). A number of shear veins with minor shear displacement were observed. Measurements were made in pieces that are oriented as such (<1% of the veins). In addition, to minimize unreliable data, veins with geometries that appeared to represent thermal contraction were excluded from measurement; these included veins with Y-shaped intersections and sinuous steeply dipping veins that are intersected by radiating veins. Most veins in Hole U1365A have planar morphology, and most irregular veins also have an overall planar trend. A number of planar veins splay at their ends. Where veining is pervasive, anastomosing veins and vein nets are common and these geometries appear to have preceded brecciation. Rare vein geometries include stepped pull-apart veins and en echelon veins. Secondary minerals within veins that indicate shear include fibrous phyllosilicates and occasionally saponite and celadonite (see “Basement alteration”). These minerals often show preferred orientation and appear to have filled the vein syntectonically (Fig. F27C). For most veins, however, shearing is not present and clay minerals grew with the long portion of the fiber orthogonal to the vein wall. In curved and irregular veins, fibers can be oblique or radiating. En echelon tension gashes indicate that shear deformation and cracking occurred during cooling of the magma.

Structural orientation

Lack of orientation of the recovered core means that only dip can be determined, with dip direction relative to an arbitrary north (see “Lithostratigraphy, igneous petrology, alteration, and structural geology” in the “Methods” chapter [Expedition 329 Scientists, 2011]). Shore-based analyses utilizing paleomagnetic data and wireline logging results will need to be carried out to reorient some of the veins. A summary of the dips and apparent strikes of veins and joints for basement at Site U1365 is shown in Figure F34. Most structures are planar features in line with the massive and thin-flow unit boundaries; however, a small number of vertical veins and joints are also present.

Structural summary

Hand-specimen observations support the following sequence of structure formation in rocks from Hole U1365A:

  1. Formation of radial cooling cracks perpendicular to pillow margins;

  2. Formation of horizontal cracks with associated hydrothermal alteration halos from which fluid flow was focused within the lava flow boundaries; and

  3. Development of younger vertical fractures without halos, which is possibly related to tectonic stresses caused by rotational movement of the basement within a fault block as the crust moved away from the spreading axis.