New ocean crust is constantly being formed from mid-ocean ridge axis. Voluminous flows of lava are emplaced also away from the ridge axis, adding off-axis crustal layers to the crustal pile. Research on deep sea is of great importance to better understand the mechanisms and the nature of the crust forming and evolution. In this work, we decipher the first stages of the postmagmatic evolution of an intact volcanic section from the upper oceanic crust at ocean drilling program/integrated ocean drilling program (ODP/IODP) Site 1256 (Eastern Pacific Ocean). Using for the first time an innovative core-log integration technique to match direct (core-related) and indirect (borehole-related) data by depth shifting and reorienting individual core pieces recovered by drilling, we are able to identify the clusterization of structures and physical properties within distinct downdeep “strong” and “weak” lava zones, reflecting the cooling and tectonic evolution of lavas rather than lithological variations. We define the evolution of the structural zones that typically affect lava flows: colonnades and entablature zones, studying an off-axis lava flow encountered in present-day upper ocean crust. For the same off-axis flow, we are also able to suggest the lava flow direction (NW-SE) and its relationships with the paleoridge axis. Despite the environmental difficulties in the study of the subseafloor under deep water and using only one-dimensional data deriving from ocean drilling, this work shows how an array of diverse data can be integrated into a coherent interpretation of lava flow history obtaining detailed information on the mechanism of submarine lava emplacement and flow.
 Many studies on faults and modes of fissure development at mid-ocean ridges have been undertaken in the last years [e.g., Macdonald et al., 1996; Buck and Poliakov, 1998; Poliakov and Buck, 1998; Buck et al., 2005; Deschamps et al., 2005]. Two main causes are invoked for crack forming in the oceanic crust, the first being tectonic fissuring associated with plate movement and intraplate deformation [e.g., Bull, 1990], the second being eruptive fissuring due to thermal contraction of volcanics and to dike intrusion [e.g., Lister, 1972, 1974; Wright et al., 1995]. Distinguishing between shear failures (faults) and Mode I failures may be relevant for understanding the nature of the failure (e.g., tectonic rather than cooling failure). Specific studies on fracture mechanics in volcanic rocks are mostly related to subaerial environment [e.g., Tomkeieff, 1940; Lister, 1974; Aydin and Degraff, 1988; Pollard and Aydin, 1988; Grossenbacher and McDuffie, 1995; Gudmundsson, 1995; Self et al., 1997; Lore et al., 2001; Kilburn, 2004; Schaefer and Kattenhorn, 2004; Lescinsky et al., 2007; Kattenhorn and Schaefer, 2008]. Such studies generally attribute the generation of vertical (column-bounding) and horizontal (column-normal) fractures to the thermal contraction of rocks and hence to their cooling history. Difficulties in operating at deep seafloor make studies on subaqueous lava less common. However, cooling-related fractures and structural zonation of lava flows, with formation of well-known polygonal colonnades, have been observed also on the seafloor [e.g., Bellon et al., 1985; Budkewitsch and Robin, 1994] or within ophiolites [e.g., Umino, 2012]. Even if, the use of submersible vehicles allows to obtain many structural informations [e.g., Auzende et al., 1996; Curewitz and Karson, 1997; Karson et al., 2002; Hayman and Karson, 2007, 2009], the best way to analyze the structure of the oceanic crust sub-basement is to drill and core it. Nevertheless, difficulties typically encountered during drilling operations in ocean (e.g., the length of recovered cores is not always equal to the length of the drilled interval [e.g., Lofts and Bristow, 1998]) may prevent a correct interpretation of the crustal lithostratigraphy. Furthermore, the absence of azimuthal orientation of cores also prevents a correct geological interpretation of the crustal structure [Fontana et al., 2010]. In order to improve our understanding of the subseafloor geological environment, it is thus crucial integrating distinct analysis approaches such as core and log measurements. This formulation may provide a more complete information and improve both thoroughness of data achieved directly on cores, and comprehension of values achieved indirectly from geophysical measurements.
 In this study, we present an integrated analysis of direct (structural, petrological, and petrophysical) data from drilled cores and in situ indirect geophysical downhole logging data to perform a structural and petrophysical characterization of the upper oceanic crust at ocean drilling program/integrated ocean drilling program (ODP/IODP) Site 1256 (Eastern Pacific Ocean). We used for the first time a technique discussed by Fontana et al.  for depth shifting and reorienting of individual core pieces recovered by drilling. Such analysis allows us to characterize the internal structure of a ca. 100 m thick lava flow encountered in the upper crust during ODP Leg 206 [Wilson et al., 2003]. The structural pattern of the lava flow is thus correctly oriented with respect to the geographic coordinates. Using only unidimensional data deriving from ocean drilling, we are able to show a qualitative solidification rate scheme, and obtaining a structural zonation, with indication of flowing direction of the studied lava portion.
2. Geological Background and Crustal Stratigraphy
 The study area (ODP/IODP Site 1256; 6°44.2′ N, 91°56.1′ W) is located on the Cocos Plate, approximately 1150 km east of the present East Pacific Rise (EPR) crest and ∼530 km north of the Cocos-Nazca Spreading Center (Figure 1). Crust at ODP/IODP Site 1256 formed 15 Ma ago during an episode of superfast spreading (full rate of ∼200–220 mm/yr) [Wilson, 1996; Müller et al., 2008]. The total crustal thickness at Site 1256, as estimated from the seismic structure, is ∼5–5.5 km [Swift et al., 2008]. Velocities of layer 2 is 4.5–5 km/s and the layer 2/3 transition is interpreted to be at 1300–1400 m sub-basement [Expedition 309/312 Scientists, 2006c]. The seafloor topography and sediment strata are well correlated with the igneous basement [Hallenborg et al., 2003].
 Four holes have been drilled at Site 1256 (Figure 2). Holes 1256A and 1256B have recovered a complete section of the sedimentary overburden, while holes 1256C and 1256D include basement penetrations. Hole 1256C reaches 88.5 m into basement and Hole 1256D is a cased reentry hole that was started during ODP Leg 206 and deepened during IODP expedition 309, expedition 312, and the recent expedition 335, to a present total depth of ∼1520 m below seafloor (mbsf) [Wilson et al., 2003; Teagle et al., 2006; Expedition 335 Scientists, 2011].
 The crustal stratigraphy at Site 1256 is shown in Figure 2. Beneath a sedimentary portion 250 m thick, the upper ocean crust at Hole 1256D was preliminarily subdivided into the following parts: a single massive lava flow (276–350 mbsf), inflated flows (350.3–533.9 mbsf), sheet and massive flows (533.9–1004.2 mbsf), a transition zone (1004.2–1060.9 mbsf), sheeted dike complex (>1060 mbsf), and plutonic complex (>1407 mbsf) [Teagle et al., 2006; Wilson et al., 2006].
 A new igneous stratigraphy of the crust drilled at Hole 1256D from 312 to 1425 mbsf has been proposed by Tominaga et al.  on the basis of geophysical properties and core-log correlation and includes the following 10 lithofacies (or “electrofacies”): massive flows, massive off-axis ponded lava, fractured massive flows, thin flows, thick pillows, pillows, fragmented flows, breccias, isolated dikes, dikes in sheeted dike complex, and gabbros (see Figure 2).
 For the purpose of the present paper, we need a simplified stratigraphy description thus, the following general subdivision into sections derived by matching the two stratigraphic interpretations is suggested:
Upper Units: the shallowest basalt portion cored at Hole 1256C from 252.4 to 280.55 mbsf; similar units have been drilled but not cored in Hole 1256D;
Lava Pond: it corresponds to the massive lava flow of Wilson et al.  and to massive off-axis ponded lava of Tominaga et al. , extending from 280.55 to 312.75 mbsf at Hole 1256C and from 276.1 to 350.3 mbsf at Hole 1256D;
Flows and Breccias: section that extends from 312.8 to 322 mbsf at Hole 1256C (the cored bottom part of Hole 1256C) and from 350.30 mbsf to the first subvertical intrusive contact that was initially located at 1004.40 mbsf in Hole 1256D [Expedition 309/312 Scientists, 2006b], even if, Tominaga et al.  identified in the logs the first evidence of isolated dike approximately 200 m above (∼816 mbsf);
Transition Zone (Hole 1256D), from the first intrusive contact down to 1060 mbsf;
Sheeted Dikes (Hole 1256D), from 1060 mbsf, down to 1406.6 mbsf;
Plutonic (Hole 1256D), from 1406.6 mbsf to the cored bottom depth of Hole 1256D at 1521.6 mbsf.
3. Structure of the Upper Crust at Site 1256
3.1. Core Structures
 The volcanic section drilled at Site 1256 is characterized by structures related to primary magmatic as well as postmagmatic processes [Wilson et al., 2003; Crispini et al., 2006; Tartarotti et al., 2006]. Primary igneous features include lava flow-related textures, laminations, aligned vesicles, folding, shear-related structures, felsic veining and patches, and igneous contacts. On the basis of structural and microstructural features observed in Hole 1256C, the Lava Pond section has been divided into five structural units [Crispini et al., 2006] which partly coincide with the magmatic units as defined by Wilson et al. . Textural variation, such as the grain size or morphology variation with depth, and distribution of varioles described by Umino  for both Hole 1256C and 1256D, appear to change in response to the distance from the chilled margin, showing the degree of undercooling, and hence, represents a key to understand the crystallization and cooling history. Postmagmatic brittle structures of the lava section at holes 1256C (252–322 mbsf) and 1256D (276–1004 mbsf) have been classified as follows by the Shipboard Scientific Party [2003b] and by Expedition 309/312 Scientists [2006b], and hereafter named “structural log”:
Joint (J): fracture with no infilling material and no evidence of displacement between the two sides. Only 75 joints were recorded and measured in cores from Hole 1256D (10 at Hole 1256C). Recognized drilling-induced fractures were not included in the structural data sets.
Microfault (F): fracture with shear displacement and no infilling material. The dislocation offset in microfaults ranges from millimetric to centimetric. 307 Microfaults were recorded in the lava section at Hole 1256D and 6 at Hole 1256C.
Vein (V): fracture filled with minerals. Veins are the most abundant brittle structures in the recovered cores. Almost all the observed pieces contain at least one vein. 3353 veins were logged in the lava section at Hole 1256D and 524 at Hole 1256C. Veins are filled with a variety of secondary minerals and their geometries are variegate: planar, irregular, splayed, etc. Thickness of the veins ranges from less than 0.1 up to 25 mm. The average value is 0.6 mm. The thickness value that most frequently occurs (mode) is 0.1 mm.
Shear Vein (SV): open fracture, mostly planar, which is coated by slickenfibers or overlapping fibers of secondary minerals that testify the slip direction. One hundred and seventy-four shear veins were detected in the lava section at Hole 1256D and 11 at Hole 1256C. Shear veins are 0.1–10 mm wide, the average thickness is 1.26 mm and the most frequent thickness is approximately 2.5 mm.
 Besides joints, microfaults, veins, and shear veins, a number of various magmatic or postmagmatic features have been described, such as the brecciated zones [Shipboard Scientific Party, 2003b; Expedition 309/312 Scientists, 2006b], yet not logged in the structural log of the present study.
 Postmagmatic structures represent conduits for fluid circulation leading to the alteration process [e.g., Curewitz and Karson, 1997] allowing the precipitations of secondary minerals (mostly clay minerals). Secondary minerals from holes 1256C and 1256D commonly fill primary igneous features, such as, voids, pores, felsic patches, felsic veins, and vesicles which often merge together and may be superimposed by postmagmatic veins (Figure 3a). In such cases, postmagmatic structures link primary igneous features by following crystal boundaries. Otherwise, postmagmatic structures are independent by igneous features and break solidified crystals (Figure 3b). Most postmagmatic structures show Y-shaped or T-shaped morphologies (with steeply dipping failures intersected by subhorizontal failures), which are typically interpreted as deriving by contractional cooling of rock [Pollard and Aydin, 1988; see, Shipboard Scientific Party, 2003b] (Figures 3c and 3d).
3.2. Downhole Structure Distribution and Orientation
 Brittle postmagmatic structures observed and measured onboard during ODP Leg 206 [Wilson et al., 2003] (see also the supporting information in Teagle et al. ) are here analyzed. Raw values of structural measurements are here processed according to the procedure described in Appendix A and taking into account the proposed stratigraphic subdivision.
 Similarly to Tartarotti et al. , we here calculated the “downhole failure intensity” (DFI), the “failure count” (FC) and the “failure density” (FD) parameters (raw and corrected number of planar structures per dm of core, respectively; see Appendix A) for every single postmagmatic structure typology in the whole lava section from holes 1256C and 1256D. We use “failure” instead of “fracture” to refer to any secondary structure that represents a crack (eventually sealed) that interrupts the cohesion of the rock.
 DFI data suggest that the Lava Pond section from both holes appears moderately failed (Figure 4). The Lava Pond from Hole 1256D (average recovery = 93%) is also much less failed than the deeper portions of the drilled lava section. Similar low-failure intensity also appears between 840 and 900 mbsf, even if this deeper portion has a low recovery (<50%).
 FC and FD are plotted versus depth for the two holes in Figures 5 and 6. Specific intervals or major spikes in FC and FD downhole distribution are here pointed out to better understand descriptions and discussion reported in paragraph 6 and 7. In Hole 1256C (Figure 5) high FD (total structures) occurs between 252 and 256 mbsf, at 260.2 mbsf, in the zone between 267.1 and 272.0 mbsf, at 275.9 and 279.6 mbsf (upper units); in the Lava Pond at 293.9; within Flows and Breccias, between 313.2 and 314.6 mbsf, at 317.5 and 322.0 mbsf. FC and FD diagrams confirm that the Lava Pond appears to be less failed then the other crustal sections. Only two spikes within the Lava Pond exceed the FD value of 3, at 281.5 mbsf (FD = 4.2) and 293.91 mbsf (FD = 5.56). The mean FD value within the Lava Pond is 0.58, whereas the mean value for the entire Hole 1256C is 1.85. In Hole 1256C, veins more than other structures contribute to the downhole failure distribution, and thus, vein spikes are quite comparable with spikes of total structures. Microfaults and shear veins occur only within the Lava Pond section. Microfaults are concentrated at 289, 294, 306, and 307 mbsf and shear veins are more abundant in the lower part of the Lava Pond section. Joints do not occur in the Upper Units; in the Lava Pond section they are scattered and are present at the top of the Flows and Breccia section, where a major FD spike occurs.
 In Hole 1256D, FC shows few peaks with a major spike at 489.2 mbsf (FC = 10; Figure 6). The Lava Pond interval has high FC but its high recovery and the high number of pieces with vertical alignment (see Appendix A) lead to a very low-FD contribution. The mean FD for the lava section at Hole 1256D is 4.80, whilst the mean value for the Lava Pond is 1.95, and for the Flows and Breccias is 5.95. The most significant FD spikes occur below 494 mbsf with the highest value (FD = 45.33) at 789.5 mbsf. The Lava Pond has the maximum FD value (FD = 7.36) at 351.3 mbsf. Other FD spikes (FD ≤ 7) occur at 276.2–276.4, 296, 303.8, 315.6, and 335.3 mbsf (Figure 6). For what concerning the various structures in Hole 1256D, veins mostly contribute to the total structures downhole plot distribution, with an exception at 742.8 mbsf where a spike in the shear veins plot is observed (Figure 6). Shear veins result scattered throughout Hole 1256D, and within the Lava Pond, they contribute to the total structure spike at 280.31 and 315.6 mbsf. Microfaults were detected only within the Lava Pond section and between 810 and 830 mbsf in the Flows and Breccias section. Joints are few and scattered throughout Hole 1256D. At 335.5 mbsf joints are the most prominent structures, and a major spike (FD = 10.7) occurs at 535.1 mbsf.
 Even if the mean values of FD from both holes 1256C and 1256D are different, the Lava Pond section results to be the least failed portion of the entire lava section, as already inferred by Tartarotti et al. .
 In addition to structure counting, the true dip angle of structures included in core pieces with vertical alignment were measured during ODP/IODP cruises [Shipboard Scientific Party, 2003b; Expedition 309/312 Scientists, 2006b]. A total of 767 postmagmatic structure dip angles in Hole 1256C and of 1790 in the lava section of Hole 1256D were measured.
 Structures from the Lava Pond and from Flows and Breccias sections at Hole 1256C mostly show gently dip angles; conversely, the true dip values in the Upper Units of the same hole are more scattered with prevalence of steeper angles (Figure 7). Mostly dip angle of all structures in the Lava Pond and in the Flows and Breccias sections are comprised between 0° and 40°. Dip angles of veins (the most abundant feature), joints, and microfaults are comparable, whereas shear veins, which were found only within the Lava Pond section, are much more steeply dipping than other structures.
 In Hole 1256D, the true dip angle of all structures gradually increases downward (Figure 8). Dip angles of Lava Pond's structures are mainly lower than 30° (Figures 8a, 8b, and 9). Structures of the upper part of the Inflated flows portion within the Flows and Breccias section (351–500 mbsf) are mainly characterized by low dips (Figures 8c and 8d) and retain a scattered attitude in the lower portion (500–550 mbsf; Figure 8e). Below 550 mbsf, postmagmatic structures are mainly steeply dipping (Figures 8f–8n).
 Despite the average dip angle of all features in the Lava Pond section at Hole 1256D is generally gentle, shear veins (mainly >80°) and microfaults (mainly ∼40°–50°) show a sharp discordant attitude being steeply dipping (Figure 9).
 In the Flows and Breccias section, more than 50% of all features have a dip angle >50° and ∼18% have a dip angle higher than 80° (Figure 9). Only joints of the Flows and Breccias section are characterized by an average gently dipping angle.
 Summing up, joints are gently dipping in both holes 1256C and 1256D and shear veins are steeply dipping in both holes 1256C and 1256D. Joints and shear veins attitude appears extremely similar in both Lava Pond and Flows and Breccias sections, and hence, these structures seem to be not controlled by depth or lithology changes.
4. Physical Properties
 We here describe shipboard analyses performed during ODP Leg 206 [Wilson et al., 2003] and IODP exp. 309 [Teagle et al., 2006] on the lava section from holes 1256C and 1256D. Detailed descriptions of procedures and methods are explained in Shipboard and Scientific Party [2003a] and Expedition 309/312 Scientists [2006a]. We used and reprocessed only a selected set of data in order to perform core-log integration (Appendix B). Original raw data are publicly available at the IODP web site (http://iodp.tamu.edu/janusweb/links/links_all.shtml).
 Limited intervals and spikes in the downhole distribution of the physical properties are here considered to better understand description and discussion within paragraph 6 and 7. Multisensor track (MST) data from Hole 1256C show very scattered values (Figure 10a). Recorded gamma ray attenuation (GRA) were extremely low in Hole 1256C and mainly lower than 2.5 g/cm3 (basalt density should be between 2.70 and 3.30 g/cm3) [e.g., Telford et al., 1990]. Every measure [GRA, magnetic susceptibility (MS) meter, natural gamma radiation (NGR) sensor] becomes less scattered below 283 mbsf (boundary between upper units and Lava Pond sections) due to the better conditions of the Lava Pond samples, mainly represented by massive, not fractured basalt. Between 289 and 293 mbsf GRA recorded low values while MS recorded a positive peak. From 294 down to 307 mbsf, NGR is high, probably due to high K concentration [Shipboard Scientific Party, 2003b]. In this segment, MST recorded low MS measurements and the highest average GRA values of Hole 1256C. Below 312 mbsf (Lava Pond–Flows and Breccias sections boundary) GRA is low while MS and NGR show high values.
 Discrete measurements of density have been considered more precise than GRA [Shipboard Scientific Party, 2003b; Expedition 309/312 Scientists, 2006b]. Bulk density (BD) from moisture and density (MAD) analysis is steady through the basement portion of Hole 1256C (average = 2.88 g/cm3; Figure 10b). Porosity and Vp show generally opposite trends through Hole 1256C (high porosity = low Vp and vice versa; Figure 10b). They are both particularly scattered in the shallowest portion of the hole above 263 mbsf, and then show a smaller range of variation down to 282 mbsf (top of the Lava Pond section) where porosity increases. At 290 mbsf discrete measurements recorded a negative peak in Vp and a positive peak in porosity. Below 307 mbsf porosity increases to higher values while Vp decreases.
 MST data for Hole 1256D are shown in Figure 11a and, similarly to Hole 1256C, display very scattered values. There is a sharp discontinuity between the Lava Pond and Flows and Breccias sections at 350 mbsf. The Lava Pond is characterized by relatively high GRA and MS and low NGR. GRA shows low values through the entire lava section of the hole (slightly higher than in the Lava Pond at Hole 1256C, but yet mainly <2.8 g/cm3) and a slight downward increasing trend, as well as for MS, while NGR slightly decreases. Between 840 and 888 mbsf there is a wide segment with relative low MS and GRA, and high NGR values. MS shows few peaks: at the top of Hole 1256D (274–289 mbsf), at 648 mbsf, and at 772 mbsf.
 BD measured on discrete samples (MAD analyses) in Hole 1256D is steady through the entire lava section (275–1004 mbsf) and possesses an average value of 2.83 g/cm3 (Figure 11b). In the Lava Pond section (275–350 mbsf), porosity is lower than 6% and Vp is mainly higher than 5000 m/s. Below 350 mbsf and to the bottom of the lava section, porosity decreases with depth. Below the Lava Pond section, there are two wide segments (350–450 and 535–688 mbsf) which are highly porous and characterized by low Vp. Segments 450–535, 688–840, and 840–1004 mbsf generally have low porosity and relatively high Vp. Approximately at 770 mbsf porosity shows a peak (∼10%). Through the deepest segment (840–1004 mbsf) porosity is always lower than 5%.
5. Indirect Observations and Analyses
 The indirect analyses presented in this chapter are based on wireline logging data collected onboard during ODP Leg 206 [Wilson et al., 2003] and IODP Expedition 309 [Teagle et al., 2006]. Standard log analyses have been undertaken in both holes 1256C and 1256D, but borehole imaging [formation microscanner(FMS) and ultrasonic borehole images (UBI)] have been performed only at Hole 1256D. The logging data have been recorded by Schlumberger in digital log interchange standard format and processed by the Borehole Research Group at the Lamont-Doherty Earth Observatory [Shipboard Scientific Party, 2003a].
 We here elaborated (discarding evident error values) and replotted all original log data which are publicly available at http://iodp.ldeo.columbia.edu/DATA/. Moreover, we performed FMS and UBI data processing (dynamic and static normalization) and we interpreted such processed data using a specific tool (Geoframe™ by Schlumberger) to identify features (relatively low or high resistivity or ultrasonic signals) and interpret them as geological structures (failures). Natural gamma ray (GR), electrical resistivity, density, photoelectric effect, neutron porosity, microresistivity and ultrasonic logs (see Appendix C) were used to assess downhole variations throughout the lava section, and correlation to FD or physical properties. Below, peculiar intervals or characteristic spikes in downhole log measurements are pointed out in order to follow the description and discussion within paragraph 6 and 7.
5.1. Downhole Log Measurements in the Lava Section at Hole 1256C
 The hole diameter detected by the hydraulic caliper at Hole 1256C (from 252 to 322 mbsf), ranges mainly between 10 and 11 in. (Figure 12a). At 256 mbsf, hole diameter results approximately 7 in. Washout was recorded at 265 mbsf (∼15 in.). A wide (>12 in.) or irregular borehole affects most recordings, particularly those that require good contact with the borehole wall (e.g., density and borehole wall images).
 Natural GR emissions slightly decrease with depth. GR shows higher values (average 7.22 gAPI) and higher variability (from 5.11 to 9.37 gAPI) in the shallowest basement portion (from 252 to 275 mbsf; Figure 12b). The deepest portion (average 5.73 gAPI) shows values between 4.13 and 7.2 gAPI (only one peak at 281 mbsf reaches 8.23 gAPI). Potassium percentage [from the spectral gamma rays data—hostile environment gamma ray sonde (HNGS)] slightly decreases, while Th increases with depth (Figures 12c and 12d). Uranium curve shows four main intervals (Figure 12e): low concentration above 255 mbsf and between 275 and 289 mbsf (mainly < 0.1 ppm); high concentration between 255 and 275 mbsf (up to 0.45 ppm), and intermediate values below 289 mbsf (mainly between 0.1 and 0.3 ppm). Either spectral gamma ray curves (K, Th, and U) show a peak at 281–282 mbsf.
 Electrical resistivity is characterized by downward increase of the recorded values (Figure 12f). There are two notable peaks at 262 and at 277 mbsf. Below 304 mbsf resistivity strongly decreases.
 Neutron porosity is characterized by a decreasing trend with three main distinctive zones (Figure 12g): in the interval above 262 mbsf, neutron porosity ranges between 17% and 45%; between 262 and 282 mbsf neutron porosity ranges between 4.4% and 27% with a peak of 53% at 264 mbsf; below 282 neutron porosity is lower than 10%.
 BD (ρb) registration shows an increasing trend with three main distinctive zones (Figure 12h): from 252 to 262 mbsf ρb ranges between 1.86 and 2.90 g/cm3; in the middle portion (from 262 to 282 mbsf) ρb ranges between 2.53 and 2.97 g/cm3 (except near the washout that occurs at 265 mbsf); in the deepest portion ρb values are mainly between 2.8 and 3 g/cm3. At 297 mbsf ρb is particularly low (2.39 g/cm3).
 Photoelectric index (Pe) measurements show an average value of 6 b/e, and generally ranging between 3.4 and 8.2 b/e with two peaks at 299.66 (14.35 b/e) and at 312.18 mbsf (18.22 b/e; Figure 12i).
 The FMS and UBI tools have not been deployed in Hole 1256C.
5.2. Downhole Log Measurements in the Lava Section at Hole 1256D
 The hole diameter values recorded by hydraulic calipers show that the hole throughout the Lava Pond section is in very good condition generally reading between 10–12 in. (∼25–30 cm) in diameter except for a single washout at ∼310 mbsf (see Figure 13a). Below the Lava Pond section (349 mbsf) and down to approximately 450 mbsf, borehole conditions are not particularly fine and care needs to be taken when working with logs in that interval. Overall, the hole diameter through the lava section is quite regular and mostly included between 10 and 12 sometimes 13 in. (see Figure 13a).
 GR data show a sharp change between the Lava Pond (275–349 mbsf; mean: 5.55 gAPI) and the juxtaposed underneath rocks from 349 to 400 mbsf (mean: 9.25 gAPI) which may have measurements being affected by poor borehole conditions (Figures 13a and 13b). The average GR value in the Lava Pond section is lower than that in the Flows and Brecias section (mean: 6.19 gAPI). The GR is determined by a linear combination of the K, Th, and U concentrations. Potassium more than Th and U contribute to the GR log, and HNGS data of K are quite comparable to the GR log (Figures 13b and 13c) especially in the shallower lava section portion. High Th count (mean 0.42 ppm) and relatively low K percentage (mean = 0.11%) occur in the Lava Pond section with respect to the Flows and Brecias section (Figures 13c and 13d; mean = 0.31 ppm and 0.14%, respectively). Low-K percentages in the Lava Pond section could be related to the absence of celadonite (K-rich mica) as secondary mineral [Alt et al., 2010]. Uranium count in the Lava Pond section does not appear significantly different from the average value in the whole lava section (Figure 13e). Besides the Lava Pond section, other few intervals show characteristic natural GR measurements: interval between 349 and 400 mbsf exhibits high GR, K percentage, U count, and low-Th count; between 400 and 474 mbsf measurements recorded several relatively thin intervals with alternated high and low GR values, mainly related to K and U variations; down to 648 mbsf curves are variable with notable Th and U peaks at 550–580 mbsf; at ∼678 mbsf there are sharp variations in GR, K, and Th measurements toward high values followed by a decreasing trend up to 770 mbsf where high K and Th values occur; below 820 mbsf average K percentage is especially low (0.04%) excluding interval between 840 and 900 mbsf where there are isolated spikes of both K percentage and U count.
 Electrical resistivity of the Lava Pond section from the dual laterolog tool (DLL) is relatively high (mean: 76–83 ohm·m) compared to the entire drilled lava section (mean: 52–56 ohm·m; Figure 13f). Below the Lava Pond section, resistivity values increase with depth toward the deepest lava section units where they are comparable with the Lava Pond values. Resistivity measurements registered between 349 and 400 mbsf are low and this can be related to bad borehole conditions. Few zones through the Flows and Brecias section show relatively high resistivity: at 445 mbsf, between 470 and 540 mbsf, at 842.20 mbsf, and from 870 mbsf to the end of the lava section (1004 mbsf; see Figure 13f).
 Neutron porosity of the lava section at Hole 1256D show very low values (6.49%) between 275 and 349 mbsf (Lava Pond section; Figure 13g). The average neutron porosity for the Flows and Brecias section is as high as 17.18%. However, in addition to the Lava Pond section, neutron porosity is particularly low also in few other zones: between 474 and 532 mbsf, 822–844 mbsf, and from 970 to 1000 mbsf.
 BD (ρb) curve of the lavas is shown in Figure 13h. Excluding the singularity at ∼310 mbsf due to a washout, the Lava Pond section records an average high ρb with low-standard deviation 0.05: minimum = 2.55g/cm3; maximum = 2.99g/cm3; mean = 2.87g/cm3. The Flows and Brecias section records a lower mean values (2.67g/cm3) with higher (0.26) standard deviation (minimum = 1.11 g/cm3, maximum = 3.30 g/cm3). Intervals at 349–450, 678, 820, 890–960 mbsf are characterized by low-ρb values probably due to a poor conditions of the borehole wall. The deepest 50 m of the lava section exhibit high-density values that are however lower than those of the Lava Pond section (Figure 13h).
Pe curve shows sharp variations at 349 mbsf (Lava Pond—Flows and Brecias boundary) and at 960 mbsf (Figure 13i). Portions above 349 mbsf and below 960 mbsf (top and bottom parts of the lava section) result in pronounced high photoelectric indexes (average 5.08 b/e and 4.26 b/e, respectively) if compared with the middle portion (average 3.65 b/e). Pe curve is characterized by a slight increase of average values with depth from ∼349 to ∼690 mbsf (where relatively low-Pe values are recorded), followed by decreasing of the average values with depth down to approximately 970 mbsf (Figure 13i).
 From FMS and UBI image analyses we were able to identify planar features that we interpret as geological structures (veins, fractures, etc.). Such data contain structural information, and hence, they are useful to integrate indirect and direct data (see next paragraph). The Lava Pond section shows good quality borehole wall images from 275 to 349 mbsf but a relative low number of interpreted features if compared with zones having similar borehole conditions (compare with 474–532 and 580–720 mbsf intervals; Figure 14). Just below the Lava Pond section down to ∼405 mbsf, FMS and UBI images show very low microresistivity and amplitude reflection, respectively. Proceeding downward, both microresistivity and amplitude reflection measurements generally increase. Downward feature distribution shows several gaps due to lack of information, which in turn are related to bad borehole conditions (see Figure 13a). Those problematic intervals (i.e., 349–405, 421–436, 450–474, 532–575 mbsf; Figure 14) have a low number of interpreted features.
6.1. Lava Pond Subdivision Based on Integrated Wireline Logging Results, Core Structures, and Physical Properties
 The structural and petrophysical analyses of the lava section highlights that the Lava Pond possesses good borehole conditions, peculiar log records, high recovery and characteristic massive rock texture. Such petrophysic characteristics allow for good correlations between core and downhole logging data, thus, the Lava Pond section is used here as a good test bed to implement a direct and indirect data integration.
Fontana et al.  showed that between indirect (downhole log measurements) data and direct (core structures and physical properties) data a depth mismatch of few meters might exist. The same authors have calculated an average depth mismatch between curated and true depth of ±66 cm in the Lava Pond section (maximum mismatch = 3.44 m). For this reason, and for the purpose of this study, in order to allow core-log integration, we processed core and log data as described in the following sections.
 The borehole image interpretation is essential for the specific methodology we use to integrate indirect and direct data [see Fontana et al., 2010]. Since borehole imager tools (FMS and UBI) have not been deployed in Hole 1256C, the core-log integration here used, and described below, has been conducted only for Hole 1256D.
 By comparing electrical resistivity, neutron porosity, density, photoelectric factor, GR, FMS, and UBI downhole signals (described in the previous paragraphs), we identified four main logging units (units A through D) and 15 subunits within the Lava Pond section in Hole 1256D, on the base of characteristic patterns of the specific log curves, different mean values for each unit, and change in the slope of the curves (see details in Appendix D; Figure 15). In the specific, structural interpretation as deduced from FMS and UBI images allowed an exhaustive subdivision of the Lava Pond into units and subunits, due to the detailed scale of investigation (generally, the used vertical scale for the structural interpretation operations was 1:8). Units or subunits are characterized by distinctive feature density, systematic downhole variation in feature dip angle (Figure 16; complete data imaging are available from the authors), and variation in dip direction (azimuth; see Appendix D; Figure 17a). We found that units (from unit A through unit D) and subunits identified from standard logs, FMS, and UBI logs are all in agreement.
 Structural data and physical properties achieved from cores in the Lava Pond section at Hole 1256D were processed for depth-shifting correction, by applying the methodology proposed by Fontana et al. . In Figure 18, logging units (i.e., units A, B, C, D as previously defined; see Figure 15) are plotted for comparison with selected depth corrected (DC) core measurements achieved by various methodologies and distinct measurements (FD, crytal size distribution, MAD porosity, MS). Such comparison shows that the subdivision criterion used for indirect measurements is still valid also for direct data on the base of characteristic patterns of the specific downhole distribution of structural and physical properties, different mean values for each unit, and change in the slope of the downhole distribution. In fact, both approaches (direct and indirect) point to the same subdivision of the Lava Pond section at Hole 1256D. Unit A shows high FD and MS, and low Crystal Size Distribution. Between Units A and B there is an abrupt change in downhole distribution for all the measured properties: FD, resistivity, and MS decrease, while Crystal Size Distribution and porosity from MAD increase. Unit C shows high-FD values, bell-shaped Crystal Size Distribution, low resistivity and relatively high porosity from MAD. In Unit C MS is sensibly lower than in the previous units. In Unit D, FD (which increases downhole), Crystal Size Distribution, and porosity distribution show low values. Resistivity is particularly high and MS does not greatly change from Unit C (for detailed subdivision description see Appendix D; Figure 18).
 Core structures corrected for depth mismatch were subsequently reoriented to the north, by applying the method illustrated by Fontana et al. . The corrected structural data from cores are plotted in rosette diagrams (Figure 17b); each plot refers to a distinct unit as previously defined by logging data. The core structures show distinct orientations among different units as already observed for structural features interpreted from FMS and UBI images (see Appendix D; Figure 17). Strike directions of “reoriented core structures” (RCS) of Unit A are mainly oriented N-S (Figure 17b), and “borehole interpreted features” (BIF) have similar, even if slightly different, strike orientation, which is mainly directed NE-SW (see Figure 17a). Average azimuth of RCS within Unit A plots into NE quadrant while BIF azimuths mainly plot to SE. However, geological structures of the Lava Pond section are generally gently dipping and small dip angle fluctuations may produce high variation in the dip direction. Similar mismatch appears also between RCS and BIF within Unit B (compare Unit B in Figure 17). Units C and D show that RCS strikes are in deep agreement with BIF strikes (Figure 17); structures in Unit C generally strike NNW-SSE, structures in Unit D generally strike NW-SE).
 Overall, RCS, and BIF data strike approximately NE-SW in the Unit A, are scattered in Unit B, yielding two main strike directions NNW-SSE and WSW-ENE in Unit C, and strike NW-SE within Unit D where they are shifted 90° respect to the shallower unit.
7.1. Structural and Petrophysical Indications From the Lava Section
 Results from the analysis of the geological structures and physical properties on both holes 1256C and 1256D allow us to interpret the structural and petrophysical behavior of the entire, so called, layer 2 of the oceanic crust at Site 1256.
 The rocks observed from the top of the basement (Upper Units, Hole 1256C) down to the bottom of the Flows and Brecias section (Hole 1256D) are all mid-ocean ridge basalts (MORBs) [e.g., Wilson et al., 2003; Teagle et al., 2006; Tominaga et al., 2009]. However, structures and physical properties show significant variations downhole.
 Dip angles of all structures increase down deep from the top to the bottom of the lava section where steep structures prevail. Dike intrusion from below may induce such verticality in the deeper structures attitude [Tartarotti et al., 2009].
 Downhole FD distribution shows that zones with high structure frequency alternate to zones with relatively low-structure frequency (see Figures 5 and 6). Such alternate variations are not linearly distributed with depth, but follow a random-like scheme. Low- or high-FD frequency zones are always associated with distinct physical properties signals (compare Figures 5 and 6 with Figures 10 through 13). Zones where FD is high (or absent due to a recovery gap) are always characterized by high porosity and low Vp, Magnetic Suceptibility, Resistivity, and BD (“weak zones”). Vice versa, low-FD values are associated with low porosity, and high, MS, Resistivity, and BD (“strong zones”).
 The Upper Units (Hole 1256C) show a FD frequency and a physical properties downhole distribution (see Figures 5, 10, and 12) comparable with that of the Flows and Brecias section (Hole 1256D; see Figures 6, 11, and 13).
 The Lava Pond yields a unique structural and petrophysical pattern. Indirect data achieved from the Lava Pond section, between 280 and 312 mbsf in Hole 1256C, and between 275 and 349 mbsf in Hole 1256D show similar petrophysical responses. The Lava Pond section, in both intervals, shows good borehole conditions, low-natural GR emissions related to a low-potassium radioactive isotope concentration, high-electrical resistivity, low-neutron porosity, high density (ρb), and matrix density (Pe). Borehole image analyses reveal a relatively low occurrence of interpreted features, in both situations, supporting the hypothesis that this crustal portion is a massive and poorly fractured single-lava flow.
 Weak and strong zones (as defined above) within the Flows and Brecias section are approximately 50–150 m thick and are not related to a single-lava flow emplacement, thus, the occurrence of weak or strong zones should derive from the postmagmatic evolution (either due to cooling or tectonic history).
 Except for the Lava Pond section, weak zones generally correspond to the occurrence of shear veins. Since shear veins represent the evidence of rock displacement, weak zones can thus be interpreted as zones of maximal crustal strain concentration. We suggest that weak zones may represent places of such preferential strain concentration.
 Overall, we are able to state that the physical properties of the lava section at Site 1256 are mainly related to postmagmatic structures, that in turn, derive from the local stress field.
7.2. Characterization of the Lava Pond Section
 The Lava Pond section retains a peculiar massive structure if compared with the whole lava section (upper units and Flows and Brecias). In this case, it may be related to a distinct magmatic evolution considering its thickness (∼70 m in Hole 1256D and ∼30 m at Hole 1256C), being emplaced as a single flow (with no or little inflation apportion) [see, e.g., Umino, 2012], and hence, with a distinct cooling history with respect to other lava flows at Site 1256.
 Assuming an average thickness of 40 m for the Lava Pond section, this would conservatively suggest an eruption volume >3 km3 [Expedition 309/312 Scientists, 2006c].
 It is likely that such a voluminous and impermeable lava body had influenced, at least locally, the hydrogeology regime of the oceanic crust. Namely, such huge volume of massive material would have sealed the pathway for fluids exchange between seawater and the underneath Flows and Brecias rocks. However, the Flows and Brecias section has registered low temperature oxidation and fixation of alkali [Alt et al., 2010], which proceeded down deep through the entire lava section. This means that the Lava Pond emplacement happened when open circulation was already active in the Flows and Brecias section.
7.2.1. Structural Characterization
 FD frequency distribution in the Lava Pond section is very low in both holes. Veins and joints generally show gentle dip angle. Instead, shear veins and microfaults yield a relative high-dip angle. Such vertical attitude may be related either to the cooling of the lithosphere [Lister, 1972, 1974], or to tectonic field implications. However, shear failures highlight relative displacement between blocks, which are related to the stress field (even if such stress may be induced by contractional cooling). Mode I failures (veins and joints) seem to be more strictly linked to the contraction by cooling rather than tectonic displacement. A kinematic analysis would help in the determination of such aspects.
 Many of the postmagmatic structures measured and analyzed in the Lava Pond section are subhorizontal and follow vesicular alignments (see Figure 3a) or crystals boundaries. Occasionally, subhorizontal structures intercept vertical or steeply dipping structures forming a T-shaped connection (see Figures 3c and 3d). Such structures are typically interpreted as cooling structures. In fact, several authors [e.g., Tomkeieff, 1940; Lister, 1974; Aydin and Degraff, 1988; Pollard and Aydin, 1988; Budkewitsch and Robin, 1994; Grossenbacher and McDuffie, 1995; Self et al., 1997; Lore et al., 2001; Schaefer and Kattenhorn, 2004; Kattenhorn and Schaefer, 2008] reported few main types of structures related to the cooling history of subaereal, as well as subaqueous, ponded lava flows. Even if specific name of cooling structure types may vary from author to author, the geometry and geological meaning of such structures do not change. According with Schaefer and Kattenhorn's  terminology, we can define some of the cooling structure types generally observed: column-bounding fractures, column-normal fractures, and Entablature fractures.
 Column-bounding fractures are generally subvertical (or normal to the top or bottom flow surface) and cause the well-known colonnade aspects with polygonal (main hexagonal) section [e.g., Aydin and Degraff, 1988; Budkewitsch and Robin, 1994], affecting both, top and bottom flow portions. Such structures, which are parallel to the drilling direction, should be rarely intersected by holes. Intersections between hole and column-bounding fractures are even more rare if column diameter (diameter of the inscribed circle of the polygonal section of the column) is large. Budkewitsch and Robin  observed that submarine basalt flows exhibit column diameters typically ∼0.2 m thick. However, they found that the cooling rate of a flow should be approximately inversely proportional to the diameter of the columns, and they suggest a solidification front that propagate at a rate of 8.4 m/year for 1 m columns, which is approximately the rate suggested by Panseri et al.  for the Lava Pond section at Site 1256. Nevertheless, column-bounding fractures are not always perfectly vertical and only few of them may be intercepted within the core (see e.g., Figure 3c and 3d).
 Column-normal fractures are subhorizontal fractures that normal intersect the column-bounding fractures (see e.g., Figure 3c and 3d) and typically propagate along vesicular layers [Schaefer and Kattenhorn, 2004] (see e.g., Figure 3a). However, some of such subhorizontal cooling-related structures may utilize other weakness zones (such as crystal boundaries) instead of vesicular alignments to propagate [Schaefer and Kattenhorn, 2004]. Column-normal fractures, being perpendicular to the hole, should be one of the most recovered core structures, and hence, they should represent many of the structures studied in this work.
 Entablature fractures are curved and irregular fractures that tie upper and lower colonnades in the internal part of the flow along a quasi-planar interface [Budkewitsch and Robin, 1994] and should affect the last solidified portion of lava [e.g., Long and Wood, 1986; Schaefer and Kattenhorn, 2004]. Geometrically, despite in early definitions [Tomkeieff, 1940] Entablature fractures are described to be vertical, they often yield low-dip angles [Lore et al., 2001; Schaefer and Kattenhorn, 2004], and in such case, Entablature fractures should be easily intercepted by the core.
 Most of the postmagmatic structures of the Lava Pond section, here described, could represent column-normal or Entablature fractures.
 Nevertheless, some failures studied in this work do not yield peculiar characteristics of cooling structures since they either lack subhorizontal dip, such as shear veins and microfaults (see Figures 6 and 8) or break already solidified crystals (typical cooling structures inherit magmatic or late magmatic weakness zones; see Figure 3b) [Schaefer and Kattenhorn, 2004]. The latter structures may be related to the hydrothermal fluid circulation that leads to hydrofracturation and may have relationships with both local tectonics or cooling. According to studies on crystallization rates measured in the Hawaiian lava lakes [see e.g., Panseri et al., 2010; Peck et al., 1977; Wright and Okamura, 1977] the solidification at the top of such lava bodies is ∼8 m/year for the first year and 1–4 m/year (in function of lava thickness) in the following years. This means that the rate of the lava solidification does not proceed in depth equally in time. We expect that such huge change in rate of solidification, between the first and the following years, should result in some measurable magmatic aspects such as, for instance, variation in FD distribution or crystal size distribution. Actually, the downhole DC-Failure distribution as well as DC-crystal size distribution in the Lava Pond section are not uniform (see Figures 5, 6, and 18). Failure frequency in RCS and feature frequency in BIF indicate that failure population tends to be clustered in the units and subunits already defined.
Panseri et al.  showed (Figure 17 of their work) an upper solidification limit for the first year that is approximately located at 280 mbsf (present-day depth). Figure 18 shows that at 280 mbsf there is a definite change in DC-crystal size distribution. The same depth corresponds to Unit A/Unit B boundary with an abrupt change in all the physical properties, geophysical measurements, and structural distribution observed during the present study and related to post magmatic aspects.
 Similar nonlinear cooling history, even if with different solidification rates, may be observed starting from the bottom boundary of lava flows and proceeding upward [e.g., Panseri et al., 2010; Peck et al., 1977; Wright and Okamura, 1977]. According to Panseri et al.'s  model, the bottom portion of the Lava Pond section should have undergone a high-solidification rate. We may expect that such high-solidification rate has produced relatively small crystals, as well supported by the crystal size distribution within Unit D of the present study (see Figure 18).
 Previous studies on cooling-related fractures spacing [e.g., Budkewitsch and Robin, 1994; Grossenbacher and McDuffie, 1995; Lore et al., 2001; Schaefer and Kattenhorn, 2004; Umino, 2012] pointed out that the smallest fracture spacing (column-bounding as well as column-normal fractures) occurs at the periphery of the basalt flow, with larger spacing in the interior portions. However, Entablature fractures are not considered in the spacing analyses [Lore et al., 2001; Schaefer and Kattenhorn, 2004], rather, they have extremely small fracture spacing.
 Summing up, we may expect the highest fracture density near the top and bottom of the Lava Pond section, where the flow surfaces were subjected to high-cooling rates and in turn characterized by a dense network of column-bounding and column-normal fractures. A region of low-fracture density should occur directly above and below the last solidified lava portion, which is the Entablature fracture zone (entablature). A high degree of fracturing should be present within the entablature.
 DC-FD in Figure 18 shows exactly the scenario discussed above. Additionally, we observe that Unit C should represent the entablature (see Figure 18), and hence, the last solidified lava portion as already inferred observing the DC-crystal size distribution.
 On the whole, we can state that magmatic/cooling evolution produced a layering in the Lava Pond section due to the melt/crust boundary migration after basalt emplacement. Such early evolution affected the late, postmagmatic deformation pattern (geological structures) of the Lava Pond section at Site 1256.
 Unit B is characterized by anomalous large phenocrysts size. Such characteristic let us suspect that an event would be interfered with a normal steady cooling history of the Lava Pond section. The likely cause, and more simplistic possibility, is represented by an inflation of new lava that changed the thermal condition of the Lava Pond section and allowed a slower crystallization with increasing of phenocrysts size. Despite the Lava Pond composition is normal-MORB, the increase of incompatible elements in this unit, as observed by Umino , is consistent with a geochemically distinct enriched MORB. Umino  states that this geochemical feature indicates a lava injection into the interior of the solidifying lava body.
7.2.2. Petrophysical Characterization
 Despite the large variation in the downhole distribution of both direct and indirect measures related to the physical properties, there is no variation in the lithology of the Lava Pond section.
 Comparison between grain-size variation and structure distribution suggests that, in the Lava Pond section, large DC-phenocryst size intervals correspond to high frequency in the DC-FD (see Figure 18). This correlation is probably due to the discretization of low-strength tensile of the rock along greater crystal boundary where stress concentration may lead to preferential crack propagation [e.g., Schaefer and Kattenhorn, 2004].
 As a matter of fact, downhole structure distribution affects the physical properties, which primary respond to the structure and related texture rather than lithology.
 There is a strict link between high DC-FD and high-rock porosity (as shown in DC-porosity from MAD and DC-FD distribution in Figure 18) or low DC-Vp, that confirms the mutual relation between failures and porosity. Magnetic properties of basalt are commonly studied in order to understand the cooling rates and states of alteration, which are expected to have an impact on the magnetic properties [Delius et al., 2003]. MS of basalt can vary widely as it is a function of concentration, composition, and microstructure (such as grain size, shape, and fabric) of magnetic minerals, which in turn depend on the cooling and alteration history of the rock [Delius et al., 2003]. Magnetic properties are affected by alteration, resulting in enhanced values of MS at the top and, even if more moderately, at the bottom zone of a typical lava flow [e.g., Bücker et al., 1998], as in the Lava Pond section (see Figure 18). Therefore, the interior of the Lava Pond section is characterized by the lowest values in DC-MS. Although magnetic properties are directly affected by particle size variations [Audunsson et al., 1992], this does not seem to totally explain the very high peak within Unit B.
 Considering indirect data, Neutron Porosity and ρb show a strict link with failure distribution. The quasi-continuous data distribution allows a more refined subdivision into subunits showing a highly porous zone in the subunit C3. Subunit C3 (see Figure 15) should then represent the lava body nucleus characterized by high concentration of Entablature fractures. The electrical resistivity of a rock formation mainly depends on the amount of water and on the pore structure geometry [e.g., Schlumberger, 1991]. Moreover, the conductivity of the alteration phases sealing cracks in MORB cannot be ignored when deriving a porosity profile from resistivity data [e.g., Pezard, 1990]. Therefore, high porosity and high FD result in low-resistivity values and vice versa (see Figure 18). GR measurements within the Lava Pond section, being influenced primarily by alteration (which in turn follows geological structures), are strictly linked to DC-FD distribution (compare Figures 15 and 18). However, inefficient convection process, due to massive character of the Lava Pond section, points to a minor variation of GR values along this portion of lava section. The Pe depends on the average atomic number of the formation. Since fluids have elements with very low-atomic numbers, they have very little influence on Pe which is a measure of the rock matrix properties. Mineralogical heterogeneity within the Lava Pond section is not so high to greatly affect the Pe downhole distribution.
 On the whole, core structures, core petrophysical data, and log data variations, among the observed units and subunits, are strictly related to primary features that are in turn linked to the solidification rate. Variation in solidification rate influences post magmatic aspects, and hence, properties modification among distinct units/subunits indicate transformations in the solidification rate.
7.3. Lava Pond Evolution Model
7.3.1. Structure's Orientation Meaning
 As already observed, orientation of structural features interpreted from borehole log images, as well as, structures measured directly on cores (after depth shifting and reorientation) show clusterization of orientation among different units (see Figure 17).
 Cooling history of lava begins when the lava is still flowing. Thus, we can expect that the dynamics of flowing and cooling history come together to affect the geological structure of the lava body. In an ideal homogeneous medium, the cracks open at angles of 45° to the direction of shearing [Jaeger, 1969] which is the flowing direction of a lava flow. In a basalt flow, cracks are expected to be subparallel to the underlying substrate, developing a steeply dipping geometry near the lava front [Kilburn, 2004]. Motion of flow favors the opening of such cracks with a dip direction in the downstream direction, nevertheless, for a two-dimensional flow, two opposite dip directions of cracks are possible [Kilburn, 2004]. Despite the verse of motion may be missinterpreted after using the azimuth of the studied structures, the direction of flow should be normal to the strike of the observed structures. Thus, from the analysis of the strike directions of the observed structures in the Lava Pond section we may infer the direction of flowing.
 Unit A should have flowed with a ∼NW-SE direction (see Figure 17). Unit B seems to have not organized orientation (see Figure 17). This is consistent with a subsequent inflation event as inferred above discussing the DC-crystal size distribution anomaly. Unit C yields two main directions of flow NNW-SSE and ENE-WSW (see Figure 17). In Unit C, which is the nucleus of the lava flow, the strain field should lead to high turbulence due to the effect of two already solid surfaces (upper and lower solidified lava portions). In Unit D, the lava would have flowed NE-SW (see Figure 17), however, it should have been subjected to the turbulence due to the effect of the solid substrate below [see e.g., Crispini et al., 2006]. In fact, even if the underlying topography is unknown, or little is known, it affects the flow direction in the lower portion of the lava body. On the contrary the uppermost part of the flow could have a more steady flow direction leaded by a kilometer scale topographic gradient.
 NW-SE direction at the top of the Lava Pond section may be strictly related to the main lava flow motion, being not affected by solid border turbulence.
 On the basis of tomographic inversion of seismic refraction data, Teagle et al.  inferred the geometry of the Lava Pond section, which has a NW-SE elongation [see also, Panseri et al., 2010] that supports a NW-SE flowing direction [see, Zucali et al., submitted]. Such direction is parallel to the magnetic anomalies transition, and hence, to the ancient EPR [Teagle et al., 2006; Tartarotti et al., 2009].
7.3.2. Cooling History
 On the basis of our results, we propose a qualitative model of cooling rate evolution for the Lava Pond section to explain its present-day structural and petrophysical characteristics.
 The lava would have flowed in a NW-SE direction as observed above. The lava cooled rapidly, especially near the periphery, where shear stress produced cracks subparallel to the substrate (Figure 19a). Vesicles and late magmatic veins migrate upward because of their relatively lower density. Where the low-viscosity fluids (vesicles) and melt (late magmatic veins) energy are too low to penetrate upward, they migrate horizontally [see, e.g., Cosgrove, 1997] into the cracks produced by shear stress [Self et al., 1997; Panseri et al., 2010].
 The external portion of the flow became solid. In the beginning, the solidification rate was high (8 m/year at the top, minor rate near the bottom), producing a fine-grained crust (Figure 19b). In this stage, column-bounding and column-normal fractures propagated toward the internal portion of the lava body and column-normal typically arranged along vesicular layers or at the crystal boundaries.
 While in the bottom portion the cooling proceeded upward with a roughly stable solidification rate for several years [see e.g., Panseri et al., 2010], at the top, after the first year, the solidification rate slowed down, the grain size increased, and FD decreased (Figure 19c).
 In our interpretation, a local upper event (inflation?) changed the thermal condition of a portion of the lava body slowing down the solidification rate with increasing of crystal size (Figure 19d).
 Toward the center of the lava body there was decreasing of fracturing due to minor thermal contrast. However, high fracturing would be present as a result of coalescing of the upper and lower column-bounding's fronts (entablature; Figure 19e). Such acme of fracturing would affect the presupposed inflated body too. Such cooling evolution would account for the unit and subunit distribution recognized inside the Lava Pond section (Figure 19e).
8. Summary and Conclusions
 Results from structural and petrophysical analysis of holes C and D at Site 1256 shows clusterization of structures and physical properties, within distinct downdeep intervals (“strong” and “weak” zones). Such clusterization reflects the cooling and tectonic evolution of lavas rather than lithological variations, in fact, the basement rocks below the sediment overburden, down to 1004 mbsf, consist of MORB only. This fact means that the physical properties, alteration pattern, and the metamorphic overprinting of the ocean crust are mainly related to structures, that in turn are the consequence of either cooling history or tectonics.
 This work (in agreement with Tartarotti et al. ) highlights that from the structural and petrophysical point of view, the Lava Pond section has a peculiar massive texture, little alteration, and characteristic physical properties. Such diversity from the deeper lava units should be related to a different emplacement condition due to an off-axis origin (which is already inferred by other authors such as Teagle et al.  and Tartarotti et al. ).
 After considering the structural and petrophysical aspects of the entire lava section, we chose the Lava Pond section for a more detailed analysis. From the refined analysis of log data it turns out that the clustering of petrophysical data also occurs within the Lava Pond section. Clustering of the downhole distribution of several log data highlights the occurrence of 4 main units and 15 subunits within the Lava Pond section. We obtained analogous results from core data (structural and petrophysical). We performed a structural study, mainly based on downhole crystal size and structure distribution, that allows us to suggest that the Lava Pond section subdivision depends on the solidification rate variations during the cooling history. At the top of the Lava Pond section (Unit A), solidification started extremely rapidly, while in the inner parts of the lava body the solidification process slowed down (Unit B). A lesser rapid solidification affected the bottom portion and proceeded upward in a quasi-steady rate (Unit D). Unit C represents the last solidified lava portion. Variation in solidification rate caused petrophysical and structural diversification within the Lava Pond section.
 We show, in this study, that, after the depth correction and reorientation processing using the method proposed by Fontana et al. , direct core data and indirect data clustering and subdivision fit each other. We are then able to identify structural and petrophysical units as specific structural zones which typically affect lava flows: colonnades zones at the top and at the bottom parts (Units A, B, and D), and entablature zone in the inner portion of the Lava Pond section (Unit C). We may also infer that a local thermal event (maybe due to a lava inflation) disturbed the normal solidification process at the top of Unit B, as observed by the occurrences of large crystals size, high FD, and anomalous MS.
 Some structures studied in the Lava Pond section, such as those that are not subhorizontal or that break already solidified crystals, do not have cooling-related origin, being related to tectonics and hydrofracturing due to the hydrothermal fluid circulation. A further analysis of such structures should be needed to define the local stress field.
 As well as down deep failure distribution, also structure attitudes, from both core data and interpreted downhole log images, show clustering of orientations within distinct units. Preferential orientation should derive from the orientation of the strain field during the flow, and hence from flowing direction, that produced geological structures subparallel to the subsurface. We suggest to use the preferential orientation of structures belonging to Unit A as the main lava flow direction indicator. In fact, Unit A has a free surface on top, that would not influence the lava flowing. We are then able to suggest that the lava flowing would be directed NW-SE, as also supported by the inferred shape of the Lava Pond [Teagle et al., 2006; Panseri et al., 2010; see, Zucali et al., submitted]. Such flowing orientation, being parallel to the paleo-EPR, supports the off-axis interpretation for the Lava Pond emplacement. Thus, we suggest in agreement with Teagle et al.  and Tartarotti et al. , that the Lava Pond section was emplaced in an off-axis basin.
 Despite the environmental difficulties in the study of the subseafloor under deep water and using only one-dimensional data deriving from ocean drilling, this work shows how an array of diverse data can be integrated into a coherent interpretation of lava flow history obtaining detailed information on the mechanism of submarine lava emplacement and flow.
 A qualitative DFI consists of a qualitative estimation of the degree of cracking based only on a visual inspection of cores. Four classes of failure intensity were established: “slight”, “moderate”, “high”, and “complete” [see, Expedition 309/312 Scientists, 2006b; Tartarotti et al., 2009].
 FC is the number of planar structures (i.e., joints, veins, shear veins, microfaults) per decimeter of core, intersecting the cut surface of the core. Only core pieces with vertical alignment reference were logged and measured. Since the thickness of the recovered and examined cores is commonly different from the complete drilled length, the FC represents a raw value that is not completely representative of the original basement. The actual thickness of the examined cores is given by the length of vertical aligned core pieces of the recovered material (generally ≠ 100%). FD represents the number of structures per decimeter of core corrected by the effects of the examined core's thickness [Tartarotti et al., 2009]. Thus, FD is the result of the following normalization:
where RO% represents the percentage of the recovered and examined core thickness over the complete cored length for any given core.
 Physical properties of basalts cored at holes 1256C and 1256D were measured directly on board in whole-cores, split-cores, and discrete samples on the MST equipped with GRA bulk densitometer, MS meter, NGR sensor, and a P wave logger. For the present study, GRA data lower than 2 g/cm3 were discarded as suggested by Shipboard Scientific Party [2003b] and Expedition 309/312 Scientists [2006b]; however, no other reductions were applied on MST data.
 Hydraulic caliper recorded hole diameter. Two gamma ray tools have been used to measure natural radioactivity in the formation: the HNGS and the scintillation gamma ray tool [Shipboard Scientific Party, 2003b; Expedition 309/312 Scientists, 2006b].
 Electrical resistivity has been achieved using the the DLL which provides two resistivity measurements having different depths of investigation: deep (LLD) and shallow (LLS) [Schlumberger, 1991]. Electricity can pass through a formation only because of conductive water (or metallic sulfide and graphite) contained in the formation itself. Therefore, resistivity logs measurements essentially depend on the water filling rock pores or absorbed in hydrated minerals.
 Neutron porosity has been measured using the accelerator porosity sonde (APS) which emits high energy neutral particles [Schlumberger, 1991]. These neutrons lose some of their energy by collision with other particles in the rock. Greater energy loss occurs when neutrons collide with an equal mass atom nucleus, e.g., hydrogen. Neutron log responds to the amount of hydrogen in the formation reflecting the amount of liquid-infilled porous, and hence, measures the rock porosity. In igneous basement, hydrogen also occur within alteration minerals such as clay minerals, therefore, neutral porosity may overestimate the rock porosity in such environment.
 Density has been measured with HLDS which consists of a radioactive source that emits medium-energy gamma rays into the formation. Gamma rays collide with electrons in the rock formation and lose some of their energy (Compton scattering effect). Count of gamma rays reaching the detector are an indication of electron density of the formation which in turn is related to the BD (ρb) [Schlumberger, 1991].
 In addition to the BD measurement, litho-density tool also measures the photoelectric index (Pe) of the rocks (gamma rays absorbed by the formation through photoelectric effect). ρb measurement responds primarily to porosity and secondarily to the rock matrix, whereas the Pe responds primarily to the rock matrix (lithology) and secondarily to porosity [Schlumberger, 1991].
 The FMS tool records four down hole continuous perpendicular electrical images, using four pads, which are pressed against the borehole wall. We generated both “static” and “dynamic” images. Static normalization is best suited for large-scale resistivity variations, whereas, dynamic normalization enhancing the contrast between adjacent resistivity levels. A similar approach was used for processing the UBI. Using a specific tool (Schlumberger Geoframe™) we have interactively identified and interpreted as planar geological structures each feature with characteristic contrasting low- or high-resistivity responses on the FMS images and characteristic contrasting low- or high-amplitude reflection on the UBI images.
AD1. Indirect Data
AD1.1. Unit A
 This unit extends from 275 to 280.53 mbsf (Figure 15) and it is subdivided into two subunits A1 and A2. Spectral gamma ray emission in subunits A1 (from 275 to 278 mbsf) presents low-K and -U concentrations, and relatively high Th. Neutron porosity is high (up to 8%). ρb decreases with depth. Resistivity was not measured in this subunit. In subunit A2 (from 278 to 280.53 mbsf), K concentration is higher than in the previous subunit (as well as U), nevertheless, it generally remains lower than 0.1%. Resistivity is relatively high and shows 81.04 ohm·m (LLD) and 74.71 ohm·m (LLS) as average values. Average photoelectric factor slightly decrease with depth (from ∼6 to ∼4 b/e). Mean ρb is 2.89 g/cm3. APS recorded low-neutron porosity values at the top of subunit A2 (5.73%). Neutron porosity increases with depth up to 7.7% at the bottom of the subunit (280.53 mbsf). Structural features of the entire Unit A from borehole image interpretation (dataset: 112) show low-dip angle, NE-SW strike direction with dip direction mainly pointing SE (Figure 17a).
AD.1.2. Unit B
 Logging Unit B is located between 280.53 and 294.43 mbsf (Figure 15). Within this logging unit there are three main subunits characterized by strong increase of resistivity with depth: subunit B1 (280.53–285 mbsf) = 57.31–54.64 ohm·m, subunit B2 (285–291 mbsf) = 85.98–81.28 ohm·m, subunit B3 (291–294.43 mbsf) = 99.94–96.01 ohm·m (average LLD and LLS values, respectively). Neutron porosity decreases from 7.53% (average values in subunit B1), through 6.84% (average values in subunit B2), to 5.62% (average values in subunit B3). Pe slightly decreases: mean value of subunit B1 is 5.82 b/e; mean value of subunit B2 is 5.59 b/e; mean value of subunit B3 is 5.46 b/e. Subunit B2 retains an average ρb = 2.90 g/cm3 while subunits B1 and B2 both recorded 2.88 g/cm3. Strike direction of interpreted features in Unit B (data set: 157) are scattered with two major directions: NW-SE and N-S (Figure 17a). Dip angles are low and generally dipping to E.
AD.1.3. Unit C
 Logging Unit C (from 294.43 to 320.20 mbsf) presents four main subunits (Figure 15) characterized by general, even not linear, slight decreasing of electrical resistivity with depth: subunit C1 (294.43–299 mbsf) = 61.69–61.18 ohm·m, subunit C2 (299–307 mbsf) = 64.16–60.10 ohm·m, subunit C3 (307–314 mbsf) = 23.39–22.76 ohm·m, subunit C4 (314–320.20 mbsf) = 55.85–52.15 ohm·m (average LLD and LLS values, respectively). Subunit C3 has the lowest average value because resistivity is here affected by poor borehole conditions (washout at ∼310 mbsf). Photoelectric factor is characterized by mainly steady average values in subunits C1 and C2 (5.16 and 5.07 b/e, respectively), sharp decreasing trend in subunit C3 and steady low-matrix density in subunit C4 (average 4.64 b/e). ρb shows relatively high values in subunits C1 and C2 (2.88 and 2.87 g/cm3, respectively) and lower density values but with increasing trend in subunits C3 (average 2.81 g/cm3) and C4 (average 2.83 g/cm3). Neutron porosity shows low values in subunits C1 and C2 (mainly < 6%), strong increasing corresponding to washout at 310 mbsf (subunit C3), and decreasing trend (even if with high values) in subunit C4 (> 6%). Natural GR emissions are slightly higher in subunits C1 and C2 (mainly due to high K = 0.11% and U = 0.79 ppm) than in subunits C3 and C4. Structural features analysis (dataset: 300) shows for Unit C two major strike directions attending NNW-SSE and ENE-WSW and mostly gentle dips for the interpreted features. Azimuths are mainly directed toward SW and NNW (Figure 17a).
AD.1.4. Unit D
 The deepest logging unit of the Lava Pond section (Unit D) occurs between 320.20 and 349.85 mbsf (Figure 15). This is the most resistive (mean LLD = 109.16 ohm·m, mean LLS = 96.62) and least porous (mean APS measure = 7.64%) portion of the Lava Pond section. Pe strongly increases with depth. Unit D is divided in six subunits with distinctive resistivity, neutron porosity, photoelectric factor, and natural radioactive emission which is related to the K concentration. Resistivity shows high values in the middle of the unit as well as neutron porosity, Pe decreases from subunit D1 to subunit D3 and strongly increase from subunit D4 to subunit D6, gamma ray related to K concentrations maximize in the middle of each subunits.
 Subunit D1 (320.20–322 mbsf): mean LLD = 101.07 ohm·m, mean LLS = 87.82 ohm·m; mean APS = 5.58%; mean Pe = 4.88 b/e; mean K concentration = 0.09%.
 Subunit D2 (322–325 mbsf): mean LLD = 113.87 ohm·m, mean LLS = 93.99 ohm·m; mean APS = 5.29%; mean Pe = 4.56 b/e; mean K concentration = 0.87%.
 Subunit D3 (325–327 mbsf): mean LLD = 97.83 ohm·m, mean LLS = 86.30 ohm·m; mean APS = 6.14%; mean Pe = 4.13 b/e; mean K concentration = 0.04%.
 Subunit D4 (327–334 mbsf): mean LLD = 137.21 ohm·m, mean LLS = 115.75 ohm·m; mean APS = 5.74%; mean Pe = 4.71 b/e; mean K concentration = 0.12%.
 Subunit D5 (334–337 mbsf): mean LLD = 108.7 ohm·m, mean LLS = 104.35 ohm·m; mean APS = 6.12%; mean Pe = 4.99 b/e; mean K concentration = 0.12%.
 Subunit D6 (337–349.85 mbsf): mean LLD = 96.30 ohm·m, mean LLS = 91.51 ohm·m; mean APS = 5.53%; mean Pe = 5.21 b/e; mean K concentration = 0.12%.
 Structural features interpreted by borehole images analysis (datasets 297) mainly strike NW-SE in the entire Unit D (Figure 17a). There are three main dip directions: SW, NE, and W. Dip angles are generally low but higher than the average dip angles of the shallower logging units.
AD.2. Direct Data
 Generally, DC-porosity from MAD, DC-crystal size distribution, and DC-FD (Figure 18) highlight a coherent subdivision of the Lava Pond section at Hole 1256D.
 DC-crystal size distribution for plagioclase and clinopyroxene phenocrysts (raw data are from Umino ) shows for the upper and bottom units (units A and D) the lowest EQD values (diameter of a circle with an equivalent area to the crystal). Also DC-FD and DC-porosity from MAD show similar low values at Units A and D (Figure 18). Strike directions of RCS of Unit A are mainly trending NNE-SSW, while azimuth (dip directions) points mainly to N-E (Figure 17b). RCS of Unit D have strikes mainly directed NW-SE and azimuth pointing to NE and SW (Figure 17b).
 Between 280.53 and 294.43 mbsf (logging Unit B; Figure 18) each DC core data measures show a “bell-shaped” downhole distribution: DC-FD, DC-porosity from MAD, and DC-crystal size distribution for plagioclase yield a maximum above 285 mbsf (boundary between logging subunits B1 and B2) while DC-MS has a peak between 285 and 290 mbsf (corresponding to subunit B2). RCS of Unit B have strike mainly directed NE-SW (a second component directed NNW-SSE is also present) and have azimuth pointing to NNW and ENE (Figure 17b).
 From 294.43 to 320.20 mbsf (logging Unit C; Figure 18) the Lava Pond section is characterized by relative high DC-FD, relative large phenocrysts of plagioclase (but not clinopyroxene). DC-MS for the same segment shows a decreasing trend down to ∼310 mbsf and a feeble increasing down to the bottom of Lava Pond section. RCS of Unit C have two major strike components NW-SE and NE-SW and scattered azimuth orientation with a major component pointing to SE (Figure 17b).
 This research used data and samples provided by the Integrated Ocean Drilling Program (IODP). Funding for this research was provided by PUR grants from the “Università degli Studi di Milano” and grants from “Dote ricerca: FSE, Regione Lombardia”. The authors are grateful to Laura Crispini for providing her samples for this research, to the Borehole Research Group (Lamont-Doherty Earth Observatory of Columbia University) for assistance during log interpretation, and to Michele Zucali for his critical comments. We are also thankful to the reviewer John Shervais whose suggestions much improve the manuscript.