The basement rocks of Site 1139 were divided into 19 pyroclastic and lava flow units (Table T6). However, because of significant alteration and faulting, many of the unit boundaries may not have primary volcanological significance. Even the identification of rock type has been difficult in several units. In the following, we describe the rocks in each unit and then, in the interpretive section that follows, we explain the rationale for our divisions. The distribution of volcaniclastic materials is summarized in Table T7.
Volcaniclastic materials in the sedimentary sequence at Site 1139 are present both as discrete tephra layers and disseminated in pelagic and hemipelagic sediments (Table T8). The foraminifer-bearing diatom-bearing nannofossil ooze in Subunit IA contains disseminated (<2%), sand-sized basaltic lithic fragments, rare small volcanic pebbles, and traces of very fine sand-sized pumice. In the foraminifer-bearing nannofossil ooze of Subunit IB, XRD analysis shows the presence of minor alkali feldspar (sanidine?) (Core 183-1139A-3R) (Table T4) (see "Lithostratigraphy"), and a few small basalt pebbles are present in Section 183-1139A-3R-1. Volcanic components in core-catcher samples show that there was a concentration of basaltic glass in Cores 183-1139A-8R and 16 R (see Table T8); these may be primary, bioturbated glassy intervals.
Three tephra horizons (intervals 183-1139A-11R-1, 107-109 cm; 38R-3, 17-19 cm; 38R-4, 80-82 cm) (Table T9) were identified in nannofossil-bearing clay and claystone of Unit II. In the gray claystone and chalk of Unit II, concentrations of volcanic clays and lithic fragments may be associated with pulses of terrigenous (hemipelagic) clay-rich sediments from late Oligocene to middle Miocene time (see "Lithostratigraphy"). Disseminated and bioturbated volcanic ash-rich domains were observed in intervals 183-1139A-18R-3, 9-10 cm, and 33R-1, 107-115 cm. Using XRD analysis, we identified alkali feldspar (sanidine?) and rare quartz distributed throughout this unit. Volcanic components were not apparent in the foraminifer nannofossil chalk (Unit III) but volcanic lithic fragments are abundant (make up ~95% of clasts) in the sandy packstone (Unit IV). Volcanic lithic and crystal (dominantly alkali feldspar) fragments make up <3% of the grainstone in Unit V.
More than 30 subrounded (Fig. F10A), massive and flow-banded, pale orange to pale green rhyolitic lava cobbles and pebbles are present at the top of basement. All of the clasts have abraded outer surfaces consistent with reworking. There are a range of internal textures from massive or planar laminated to clasts with convoluted flow banding and a few breccia pebbles. However, planar flow-banded lava clasts are the most common with less and approximately equal amounts of convoluted flow-banded and massive cobbles (Fig. F10B). The average clast size is ~3 mm× 5 mm, and pebbles are elongate with a short axis to long axis ratio of ~70%. Most clasts have feldspar phenocrysts present, and, in the more massive cobbles, these phenocrysts are up to 2 mm in diameter. There are one or two dark gray to pink and white, altered (silicified), possibly mafic lithologies represented, but these form a minor component.
This 47-cm-thick interval of bioclastic sandstone is very similar to the grainstone (Unit V) described in the lithologic section at Site 1139 (Figs. F7, F8). In thin section, there are <1% crystals, volcanic lithics, and glass shards in this interval (see "Lithostratigraphy").
A 44-cm interval of altered, dominantly closed framework breccia, with white to pale green partially altered clasts, is best represented at the base of Section 183-1139A-52R-1 (Fig. F11). At the top of Section 183-1139A-53R-1, we find abraded rubble and a breccia cobble, a massive green flow-banded cobble, and a wholly silicified dark colored cobble that appear unrelated to the rest of Subunit 1C. The pumice/flow-banded felsic breccia has an average clast size of 1.0 cm × 1.7 cm and a maximum clast size of 2.5 cm × 5.0 cm. In thin section (interval 183-1139A-52R-1, 120-123 cm), the clasts are banded perlite. The spheroidal perlitic fractures reflect quenching and hydration of felsic glass and the alignment of perlite kernels suggests that these glassy clasts were originally flow banded (Fig. F12). Some clasts in the breccia resemble pumice, but pumice was not observed in thin section. The matrix contains coarse sand-sized fragments of similar composition to the clasts in a clay-sized matrix. The matrix is more indurated than the clasts.
This interval is dominated by dark orange to red, altered, massive perlitic glass (Fig. F13) with <10% lithic clasts. The average lithic clast size is 1.3 cm × 2.0 cm and the maximum clast size is 2.5 cm × 3.5 cm. Spheroidal perlitic fracture sets, from 0.5 to 1.5 cm in diameter, are present in the glass and are best represented in interval 183-1139A-53R-1, 110-122 cm. In thin section (Sample 183-1139A-53R-1, 127-130 cm), excellent examples of perlitic fracture textures are preserved, including nested perlite kernels (Fig. F14) and evolution of longitudinal and secondary fractures. There are also flow textures in the glass and rotation of included clasts (Fig. F15). One fragment of phenocryst-rich banded perlite is incorporated in the thin section (Sample 183-1139A-56R-3, 93-97 cm) and is a flow-banded glass clast (Fig. F16).
A subhorizontal fracture set displaces the perlitic glass of interval 183-1139A-53R-2, 9-23 cm (Fig. F17). Near one large, relatively open fracture, there is more abundant clay-mineral formation around perlite kernels (Fig. F17). Toward the base of Subunit 1D, the disrupted perlitic glass is gradational into breccia with an indurated, massive, light-colored matrix. In the interval 183-1139A-53R-2, 38-127 cm, felsic breccia with variably altered clasts of perlitic glass and incorporated lithic fragments is the principal lithology (Fig. F18).
Resinous to powdery, green and red (oxidized), clay-rich altered rock in interval 183-1139A-54R-1, 22-56 cm, has been intensely sheared. Below this (interval 183-1139A-54R-1, 56-108 cm) is a suite of variably sheared, dark and pale green, altered clastic rocks with subhorizontal fabric delineated by pale green wispy to angular granule-sized clasts. The preserved texture contains primary pyroclastic features, but alteration obscures the primary textures of the clasts (see "Alteration and Weathering"). At the top of Section 183-1139A-54R-1 (interval 183-1139A-54R-1, 0-22 cm), there is an assortment of pebbles that appear similar to cobbles in Subunit 1A. These have probably fallen downhole during drilling. The uppermost piece (interval 183-1139A-54R-1, 0-10 cm) is a massive pale orange rhyolite. The second piece (interval 183-1139A-54R-1, 10-15 cm) is a massive bright orange pebble with incorporated lithic fragments and an abraded outer surface. In interval 183-1139A-54R-1, 15-22 cm, massive orange-brown rubble has little preserved internal texture. It is possible these pebbles are all derived from the same place as the collection of rounded cobbles observed in Subunit 1A.
Unit 2 consists of dark red (oxidized) welded vesicular rhyolite with abundant (<20%), commonly broken, sanidine and minor quartz crystals (Sections 183-1139A-55R-1, 16-72 cm, and 56R-1, 11-78 cm). Flattening and agglutination textures are common (Fig. F19). Domains of flattened wispy clasts define a subhorizontal fabric in clasts with subparallel trains of void spaces (vesicles?) enhancing the texture. These rocks are very similar to the uppermost material in Unit 4. In thin section, the clear banded texture may reflect the flattening of clasts. No glass shards can be recognized; however, the sample is highly altered (Fig. F20). The lowermost part (interval 183-1139A-56R-1, 68-78 cm) is brecciated with matrix-supported clasts (0.2 cm × 0.3 cm) near the base; it has a more clastic texture than the overlying welded rhyolite and becomes green near the contact with Unit 3.
At the top of the Section 183-1139A-55R-1, there are two gray to pink and white rounded cobbles that look similar to those described as a wholly silicified (mafic?) lithology from Subunit 1A and one small altered white pebble of unknown origin. At the top of Section 183-1139A-56R-1, there are two small, green, silicified breccia pebbles with pale orange flow-banded clasts, which appear to be unrelated to the Unit 2 rocks. These rocks may have fallen downhole during drilling.
The contact between Unit 2 and Unit 3 (interval 183-1139A-56R-1, 78-80 cm) is sheared, and color changes from dark red to green across the contact. Although the crystal composition (sanidine and minor quartz) and abundance (<20%) remains similar across the contact, the internal texture of the rock changes significantly. The green crystal vitric tuff-breccia is highly altered to green clay, is largely unconsolidated, and has been disturbed by drilling at the top of Sections 183-1139A-56R-2 and 56R-3. The material can be crumbled and reduced to individual crystals and relict perlite kernels in a clay matrix. There are coherent domains in this interval that retain banded textures (e.g., intervals 183-1139A-56R-1, 126-133 cm; 56R-2, 50-56 cm; and 56R-3, 68-76 cm) (Fig. F21) and some subangular lithic pebbles (e.g., interval 183-1139A-56R-3, 117-119 cm). The lower part of Section 183-1139A-56R-3 (76-150 cm) is more consolidated, and a thin section from this part of the core (Sample 183-1139A-56R-3, 94-98 cm) shows ~15% sanidine and ~4% quartz suspended in relict perlitic glass (~40%) with clay minerals (~40%) in the interstices between perlite kernels (Fig. F22). At the base of Section 183-1139A-56R-3 (130-150 cm), there is a faint subhorizontal fabric within the tuff, which is not related to an enclosed clast.
In Section 183-1139A-57R-1, the green color continues through a finer grained domain (interval 183-1139A-57R-1, 8-45 cm), which has an oblique contact with the more crystal-rich tuff in Piece 2 and has a basal breccia in Piece 7. This finer grained material is massive and may be either a large clast or a finer grained interval lower in the stratigraphy of the tuff. Beneath this interval is a domain (interval 183-1139A-57R-1, 45-92 cm) of dark green material with up to granule-sized pale green clasts (pumice?) and a similar proportion of crystals (<20%) as the crystal-vitric tuff. In parts, this interval is red (oxidized), slightly brecciated, and has contacts with enclosed clasts (e.g., interval 183-1139A-57R-1, 84-92 cm). The contact with the underlying Unit 4 is defined by a strong color change from dark green to dark red.
Unit 4 consists of dark red (oxidized) rock that is moderately vesicular, crystal rich, and massive. In these regards it is very similar in texture to Unit 2. In detail, the pieces show a variety of textures and two different styles of brecciation. Even relatively massive intervals contain clasts up to 10 cm long, with concentrations of elongated vesicles along the sutures (Fig. F23). Many of the vesicle-rich zones are subvertical. The rock contains ~5 vol% equant to elongate, very angular, 0.3- to 5-mm vesicles, though vesicularities as high as 15% are present locally. Only interval 183-1139A-60R-2, 60-127 cm, is massive with no evidence of brecciation. Phenocrysts are present and randomly oriented.
In the upper part of Unit 4, the breccia has 0.5- to 2-cm clasts with irregular shapes cemented to each other without filling voids and vesicles. Voids make up <50% of this type of breccia, giving it a highly vesicular to pumiceous appearance. A second breccia, with <1.5-cm angular fragments, commonly with jigsaw-fit textures, contains ~40% clay-rich matrix, makes up ~10% of Sections 183-1139A-57R-1, 57R-2, and 57R-3, and dominates Sections 183-1139A-59R-1, 60R-1, and 60R-2 (Fig. F24).
The 81 cm of rock recovered from Unit 5 consists of a highly fractured and brecciated moderately feldspar-phyric trachyte with pink, green, and white alteration (Fig. F25). The clasts and more coherent parts of the unit are massive and featureless, except for randomly oriented feldspar phenocrysts. Core 183-1139A-61R-1, 0-33 cm, consists of a breccia with >1-cm subangular, subequant clasts making up 50-60 vol%. Only 15-20 vol% of the breccia is <1 mm in size, mostly consisting of a reddish black (5R 2/2) clay. The clasts between 1 mm and 1 cm in size are mostly subrounded. The underlying rock appears to have been coherent with variable degrees of fracturing. The massive nature of the rock makes it difficult to assess the proportion of distinct clasts and the amount of clast rotation in the fractured zones. Many zones do not appear to be breccias related to the emplacement of the rocks of Unit 5; they have a cataclastic appearance. The bottom of the unit has a highly altered breccia with some matrix-supported textures.
The 6.72 m of aphyric
trachybasalt recovered from Unit 6 consist of one core of highly altered and
disturbed breccia and a second core of more coherent pieces. Core 183-1139A-62R
consists of a very dark red to reddish black breccia with a clay-rich matrix
(Fig. F26). The rock fragments
within the clay matrix are angular, with fractures cutting through the smaller
(more altered) clasts but generally circumventing the larger (~10 cm)
less-altered clasts. The angular rock fragments and more coherent pieces of rock
are from a volcanic breccia dominated by subangular-subrounded clasts of mafic
lava. The lava has ~1-5 vol% vesicularity with 0.1- to 4-mm elongate, subrounded
vesicles. Core 183-1139A-63R was also poorly recovered but the 5- to 30-cm-long
pieces of core are mostly coherent lava petrographically similar to the clasts
in Core 183-1139A-62R. However, the rocks in Core 183-1139A-63R are denser with 
1
vol% vesicularity and a bimodal vesicle size distribution (<1-mm round and
>1-cm elongated vesicles). Core 183-1139A-63R-2 also contains two 30-cm-long
intervals of breccia with a clay-rich matrix. This breccia has a cataclastic
morphology and the clay matrix contains abundant well-formed slickenslides.
The 2.72 m of aphyric
basaltic trachyandesite recovered from Unit 7 consist of 1.88 m of coherent lava
sandwiched between highly altered breccias. The top of Unit 7 (Core
183-1139A-64R-1, 0-46 cm) is a dusky red (5R 3/4) to reddish black (5R 2/2)
breccia. The dominant clasts have horizontally elongated fluidal shapes with
fine-grained margins along contacts with clasts of different lithologies. In
some cases, the fluidal clasts envelop a medium-sand-sized, feldspar-rich sand.
Vesicularity of the clasts varies from 1 to 20 vol% with elongated to round, 1-mm
vesicles. At Sample 183-1139A-64R-1 (Piece 9, 47 cm), there is an abrupt switch
to a light gray coherent lava. Section 183-1139A-64R-1 (Pieces 9-12) contains 7%
irregular <0.1- to 3-mm vesicles. Beneath these pieces to Section
183-1139A-64R-2, 80 cm, the lava generally has 3-5 vol% horizontally elongated
ellipsoidal 2- to 5-mm-long vesicles. There are two exceptions. At interval
183-1139A-64R-1, 73-77 cm, a 1 cm × 3 cm rectangular vesicular domain is
incorporated in the denser lava, and at interval 183-1139A-64R-1, 124-126 cm, a
1.5 cm × 2 cm megavesicle is found. Below Section 183-1139A-64R-2, 80 cm, the
lava becomes much more vesicular with 15-20 vol% angular irregular 0.5-mm to
2-cm vesicles. Also, in the working half, at Section 183-1139A-64R-2 (Piece 4A,
93.5 cm) a 3 mm × 3 mm xenolith containing larger crystals is seen. The last
coherent rock (interval 183-1139A-64R-2, 101-104 cm) has fewer and smaller
vesicles. The remainder of Unit 7 is a sheared and altered breccia with pale red
(5R 6/2) clasts in a light greenish gray (5GY 8/1) clay.
The 3.33 m of aphyric trachybasalt recovered from Unit 8 consists of a highly altered, sheared, and disturbed breccia overlying a more coherent lava. The breccia has black (N1) clasts with a pale blue (5B 6/2) clay matrix. Common slickenslides are found in the clay. The clasts up to 2 cm in size crumble in the hand, and original features are difficult to discern. The larger clasts have 5-7 vol% elongated round, <1-mm-diameter vesicles. The lava is less brecciated and altered (but still highly fractured) below Section 183-1139A-64R-3, 77 cm, except for interval 183-1139A-64R-4, 5-21 cm, with recovered pebble-sized loose pieces with >3-cm-diameter pancake-shaped vesicles. The lava from interval 183-1139A-64R-3, 77 cm, to 64R-3, 133 cm, has highly variable vesicularity with domains of uniform vesicularity defining horizontally elongated zones >5 cm long and 0.8-2 cm thick. The vesicular parts contain 20-50 vol% very irregular 0.5- to 15-mm vesicles, but the denser parts contain 1-2 vol% angular elongate <1- to 5-mm vesicles. The lava also contains large rounded "xenocrystic" feldspars (see "Igneous Petrology"). From Section 183-1139A-64R-3, 133 cm, to the rubbly zone noted earlier (interval 183-1139A-64R-4, 5-21 cm), the lava is more uniform with ~3 vol% horizontally elongated generally 1- to 3-mm vesicles (Fig. F27). Vesicle size increases downward until the zone of low recovery. The first lava below this rubbly zone contains 2.5- to 5-cm horizontally elongated rounded vesicles. The remaining lava has 5%-15% inclined subrounded vesicles 0.5-15 mm in size with 3-cm vesicular clots. Vesicle size decreases and sphericity increases with depth. Vesicularity decreases to 2-5 vol% below Section 183-1139A-64R-4, 83 cm, with horizontally elongated ellipsoidal 0.5-to 20-mm vesicles. Inclined wispy mesostasis blebs also appear. Vesicularity increases at the base of the unit in Section 183-1139A-64R-5, from 3 vol% at 10-15 cm to 5 vol% at 21-29 cm, and reaching 7-10 vol% at 30-40 cm. Vesicle size also increases from 0.3-2 mm to 0.2-8 mm to 0.5-30 mm over this same interval. The last centimeter of Core 183-1139A-64R (and Unit 8) is altered to a dark gray.
The 1.61 m of aphyric trachybasalt recovered from Unit 9 is mostly coherent lava. The top 10 cm is a single piece of breccia with 1.5- to 4-cm rounded, folded clasts defined by vesicular domains. Vesicularity ranges from 5%-15% with irregular and elongate ellipsoidal vesicles 0.5-6 mm in size. The clasts have a uniform groundmass texture and are cemented to each other with no matrix. The underlying coherent lava exhibits a varied range of vesicularity and vesicle shapes (Fig. F28). The lowermost few centimeters of the coherent part of Unit 9 contain relatively few and small vesicles.
The 2.42 m of sparsely
plagioclase-phyric trachybasalt recovered from Unit 10 is generally coherent,
and the distribution of vesicles is plotted in Figure F29.
The upper part of Unit 10 (Core 183-1139A-65R-2, 46-110 cm) consists of loose
rounded pieces of altered lava. The uppermost pieces have fluidal shapes, 20%
elongated to round vesicles (Fig. F29).
Farther down, denser lava with ~3% coalesced round vesicles dominates. The black
(N1) clasts are partially altered and are surrounded by greenish gray (5G 5/1)
clays. The first piece of coherent lava is in Section 183-1139A-65R-2, 110 cm.
The remainder of Section 183-1139A-65R-2 consists of irregular patchy 3- to
10-cm subcircular vesicular domains. Vesicularity varies from 5 to 30 vol% with
smaller (0.1-0.3 mm vs. 0.5-4 mm) vesicles in the denser parts. Vesicle shapes
are irregular and coalescing. Below these domains only loose pebble-sized rocks
were recovered until Section 183-1139A-65R-3 (Piece 2), which contains a
boundary between a vesicular lava with mesostasis blebs and a 5-cm-thick lower
less-vesicular part that is darker. Below this dense zone, the lava becomes
highly (
20%)
vesicular with irregular vesicles composed of coalescing round vesicles. This
vesicular zone changes to less vesicular (1
vol%) lava over the interval 183-1139A-65R-3, 47-51 cm, (Fig. F30);
this lava contains wispy mesostasis blebs in Core 183-1139A-65R-3, 82-85 cm.
Vesicularity increases again at Section 183-1139A-65R-3, 90 cm, and subtle
zonations in vesicularity zones become convoluted in interval 183-1139A-65R-3,
90-97 cm. A fine-grained zone surrounds a clast at 95-96 cm. A sharp
subhorizontal, subplanar transition at interval 183-1139A-65R-3, 104-105 cm,
contains lava with a high density of very small vesicles overlying lava with
high vesicularity and larger vesicles (Fig. F29).
A 0.5- to 1-cm vein follows this transition, obscuring the original
millimeter-scale features. The vesicles in the underlying lava gradationally
increase in size until reaching a sharp contact at Section 183-1139A-65R-3, 124
cm (Fig. F31). Below this
contact the lava is nonvesicular until Sample 183-1139A-65R-3, 131.5-136 cm,
where vesicularity increases and the vesicles show a strong fining downward
sequence. At Core 183-1139A-65R-3, 136 cm, a 1-cm-thick subplanar dense zone
with very small vesicles is present before vesicularity increases, then
decreases again, reaching yet another subplanar dense zone with very small
vesicles in Section 183-1139A-65R-3, 144 cm (Fig. F32).
This lowermost zone partially envelops two 1- to 2-cm breccia clasts underneath
it.
The 3.26 m of aphyric trachybasalt recovered from Unit 11 consists of coherent lava sandwiched between two breccias. The uppermost part of Unit 11 (Section 183-1139A-65R-4) is poorly recovered and consists of loose pieces of breccia and a 10-cm piece of coherent lava. The breccia contains ~40% reddish black (5R 2/2) clasts in an ~60% dusky blue-green (5BG 3/2) clay matrix. The clasts contain 15%-20% irregular rounded vesicles and the coherent piece below contains ~3% angular vesicles. Section 183-1139A-65R-5 contains 135 cm of intact, well-preserved lava. The top of the lava has a breccia with 0.5- to 4-cm angular clasts containing 10% small irregular angular vesicles and a matrix of lava containing 15%-20% larger, extremely irregular and angular vesicles. The margins of the clasts are gradational and are marked by a zone of the matrix with decreased vesicularity (5-10 vol%). Mixing of denser and more vesicular lava continues through all of Section 183-1139A-65R-5 with dense (less vesicular) lava becoming more dominant at depth. The vesicular clast at Section 183-1139A-65R-5, 50 cm, is striking with a 2-cm vesicle or void attached to its upper surface (Fig. F33). The lava becomes more coherent in the upper part of Section 183-1139A-66R-1, with a gradual increase in vesicularity. Vesicular domains 1-1.5 cm in size reappear at interval 183-1139A-66R-1, 92-102 cm, immediately above the transition to a breccia with large (now carbonate- and zeolite-filled) voids. The breccia clasts have the classical jagged, gnarled, irregular shapes of aa clinker. The remainder of Unit 11 consists of loose altered rubble. Identifiable chunks within the rubble appear to be a breccia with similar vesicularity as the well-preserved aa clinker.
The 2.92 m of aphyric trachybasalt recovered from Unit 12 consists of an aphyric basalt with a heavily altered and brecciated top, massive interior, and a relatively well-preserved base. Only the top 50 cm of Unit 12 is a breccia, which exhibits a reddish black alteration and has a bluish gray clay matrix. The identifiable clasts have 25%-30% irregular but rounded vesicles 1-3 mm in size. These clasts are 0.5- to 1-cm angular fragments. The coherent lava underneath contains distinct clasts that are 1-3 cm in size and contain ~20% irregular and elongated vesicles 0.1-15 mm long. Within the coherent lava, more vesicular and denser parts alternate on a 10-cm scale. The lava becomes dense below Section 183-1139A-66R-2, 130 cm, with 3%-5% irregular elongated vesicles 0.3-10 mm in size. This massive lava also contains scattered megavesicles up to 6 cm in size and a highly vesicular zone in interval 183-1139A-66R-3, 24-34 cm. This interval contains 25%-30% highly irregular vesicles 0.1-6 mm in size. Near the base of Unit 12, vesicularity, fracturing, and alteration all increase. In interval 183-1139A-66R-3, 100-115 cm, vesicularity increases to 5%-7% and the subrounded vesicles are elongated horizontally. At the base of this vesicular zone, the groundmass becomes darker.
The most stunning feature in Unit 12 is a colorful package of sediments with contorted laminations (Fig. F34). The sediments include a layer of red clays, brown silty material, and green fine to very fine feldspar-rich sand. Beneath these sediments is a 15-cm-thick coherent lobe of lava with dark (chilled?) margins and 20%-25% large rounded vesicles that are elongated parallel to the margins of the lobe. Below this lobe, the lava consists of 1- to 5-cm fragments with dark (chilled?) margins. The margins themselves are fragmented and contain 3%-5% small round vesicles.
The 3.25 m of aphyric trachybasalt recovered from Unit 13 consists of a highly altered and disturbed brecciated top, a massive interior, and a poorly recovered fractured basal section. The breccia at the top of the unit extends through all of interval 183-1139A-66R-4 to 66R-5, 4 cm. The breccia consists of a mix of reddish black angular vesicular lava fragments with a matrix of bluish clays and white carbonates. The first moderately coherent lava is a 16-cm piece with vesicular margins where the vesicles are elongated parallel to the margins. This lava contains 5%-10% irregular angular vesicles 0.1-7 mm in size. Below this piece is another 10 cm of breccia with 2- to 5-cm equant rounded clasts. Very dark red (5R 2/6) 1- to 5-mm angular fragments surround these clasts. The clasts contain 3%-5%, generally <0.2-mm vesicles, and a few 2- to 8-mm irregular cavities. Below this breccia, the lava is coherent with the same characteristics as the 16-cm piece above. The vesicularity in the coherent lava varies on a 6- to 8-cm scale until Section 183-1139A-66R-5, 63 cm. In this variable upper part of the flow, vesicularity ranges from 3%-15% irregular rounded to subangular vesicles 0.1-10 mm in size. Below this, vesicularity and vesicle size continually and gradationally decrease, but vesicle shapes become more spherical. All of Section 183-1139A-66R-6 and the upper 16 cm of Section 183-1139A-66R-7 have 0.5%-1% vesicles and faint wispy mesostasis blebs. The lowermost 25 cm of Unit 13 is heavily fractured and is largely altered to a grayish red (5R 4/2) clay. The lava fragments contain 3%-7% irregular subhorizontally elongated vesicles 0.2-20 mm in size.
The 7.23 m of recovered rock from Unit 14 is basaltic trachyandesite with a complex, altered, and disturbed breccia overlying a more massive interior. From the top of Unit 14 until Section 183-1139A-67R-3, the breccia is highly altered and fractured and was recovered as rubble. The clasts are generally reddish black and the matrix includes a bluish clay. Intact clasts range from 3 to >15 cm in size and have preserved irregular fluidal shapes. Smaller clasts are not readily recognized amid the heavy alteration. The better preserved clasts have denser rims with smaller vesicles elongated parallel to the clast margins. Vesicularity in these clasts ranges from 7 to 20 vol%, from spherical to highly elongated and subrounded, and sizes from 0.1 to 5 mm. Clasts with irregular angular vesicles are notably absent. In Section 183-1139A-67R-1, 40 cm, an autolithic clast is present (Fig. F35), and, also in this section, clasts are commonly fused to each other without visible secondary cementation.
Below Section 183-1139A-67R-3, 70 cm, the volcanic breccia is no longer heavily fractured or altered and primary textures are more readily recognized. In this region, an entire continuum of fragment sizes is visible, ranging from fine sand to >6-cm cobbles. The larger clasts are generally equant and subangular to rounded but clasts <1 cm are sometimes angular (Fig. F36). In the upper part of the breccia, the clasts >1 cm in size make up 70-80 vol%, whereas sand-sized fragments make up 10-20 vol%. With depth, the alteration decreases, the fine-grained component becomes volumetrically more important, and coherence of the breccia increases. By interval 183-1139A-67R-4, 20-30 cm, clasts >1 cm in size make up only 10-20 vol%, and the fine-grained component has become a coherent lava. At this point the clast margins are also indistinct and are more defined as vesicular patches.
Interval 183-1139A-67R-4, 51-104 cm, has a different style of brecciation. The breccia consists of 0.5- to 15-mm very angular, subequant clasts of a single lithology. This region is heavily altered but contains 3%-5% irregular rounded vesicles 0.1-5 mm in size.
Below this breccia, the lava is massive with generally 0.1-0.5 vol% vesicularity. Inclined mesostasis blebs are common until Section 183-1139A-67R-5, at 50 cm. In the bottom 20 cm of the zone with mesostasis blebs, the blebs are associated with 0.1- to 0.5-mm irregular vesicles. Within the massive lava is a single 3-cm-diameter vesicular clot (interval 183-1139A-67R-5, 133-136 cm) with 5 vol% spherical 0.1- to 1-mm-diameter vesicles.
The lava becomes more vesicular near the base, starting in Section 183-1139A-67R-5, at 104 cm. Vesicularity increases as centimeter-scale indistinct vesicular pods. By interval 183-1139A-67R-5, 127-138 cm, the lava is dominated by the vesicular lava, and the dense lava makes up only 1-4 cm indistinct pods. The dense lava has identical character to the massive interior, whereas the vesicular lava has 7%-10% highly irregular vesicles 0.1-4 mm in size. Below this, the lava is highly fractured and altered. The lava appears to contain 0.5%-3% irregular angular vesicles 0.1-3 mm in size, and mesostasis blebs are common.
The recovered 4.16 m of Unit 15 is a sparsely plagioclase-phyric basaltic trachyandesite flow with a massive interior and a small basal vesicular zone. The top of Unit 15 is very highly altered, fractured, and disturbed, making it impossible to identify primary volcanic features. The degree of alteration and fracturing decreases with depth, and the first recognizable pieces of lava appear at Section 183-1139A-68R-2, 13 cm. This lava contains 5%-10% highly irregular, subrounded vesicles 0.1-15 mm in size. Within the fractured zone, there is no evidence for clast rotation, chill margins on clasts, or mixing of lava types. Over the interval 183-1139A-68R-2, 50-90 cm, the irregular vesicles become less common and a second population of more spherical vesicles appears. At the end of this interval, the lava contains only 1%-3% moderately spherical and highly rounded vesicles 2-4 mm in size. A high degree of fracturing hides most primary features in the lava below this level, but, in general, the lava becomes dense with a 0.5-m-long vesicular zone from Section 183-1139A-68R-3, 86 cm, to 68R-4, 16 cm. In the upper massive part, vesicularity appears to be <0.5 vol%, and a megavesicle and inclined mesostasis blebs are tentatively identified. The vesicular zone consistently contains 5-10 vol% vesicles, and vesicle sizes range from 0.1-25 mm. However, vesicle shape changes dramatically with depth. At the top, the vesicles are highly irregular and subangular. In the middle, the vesicles are rounded and elongated, in the vertical direction higher up and horizontally lower down. Toward the bottom of the vesicular zone, the vesicles are once again subangular and elongated in an inclined or subhorizontal direction. These vesicles are interleaved with, and grade into, faint wispy mesostasis blebs. The lower massive portion of the flow contains 0.5%-1% near spherical vesicles 0.5-1.5 mm in size.
The base of the flow grades into the basal vesicular zone. Vesicularity begins to increase in Section 183-1139A-68R-4, 77 cm, with the appearance of vesicular clots a few centimeters across. These become more dominant with depth so that in Section 183-1139A-68R-4, by 100 cm, the lava is entirely made up of coherent vesicular lava. In this basal vesicular zone, the "dense" lava actually contains 10%-15% small, extremely irregular and angular vesicles, whereas the vesicular clasts/lava contain 10% rounded vesicles, 1-6 mm in size.
The 2.85 m of recovered aphyric basaltic trachyandesite from Unit 16 contains a breccia overlying a massive interior. The upper part of the breccia (interval 183-1139A-68R-4, 100-145 cm) is too heavily altered, fractured, and disturbed to identify any features of the original lava. The lava is better preserved but brecciated down to near the base of Section 183-1139A-68R-5, 140 cm. This breccia shows a range of clast sizes with clasts >1 cm making up 30-40 vol%, clasts 1 cm-1 mm in size making up 40-50 vol%, and fines <1 mm making up 15-20 vol%. There is no visible size sorting in the well-preserved part of the breccia. The larger clasts are generally nearly equant and subangular. The <1-cm clasts are generally angular and can commonly be jigsaw fit to larger clasts with fine-grained alteration material filling the interstices. The clasts of lava exhibit a range of shapes with vesicularities ranging from <1% to 20%. In rare cases, the larger clasts have elongated vesicles parallel to the clast margins and more spherical vesicles in the center. The one clast that stands out extends across interval 183-1139A-68R-5, 100-116 cm, and consists of dense lava. The first coherent lava at the top of the massive interior of the flow has identical vesicularity. The top of the coherent lava also has contorted banding defined by stretched and contorted vesicles that are highlighted by alteration (Fig. F37). The massive lava that makes up the rest of Unit 16 has 5%-10% microscopic vesicles (<0.1 mm) and <1% macroscopic (0.5-3 mm) highly irregular vesicles. The lava is highly altered and fractured, but it appears that the lowermost 5 cm of Unit 16 is slightly more vesicular with 1%-2% irregular 0.1- to 0.5-mm vesicles.
The 5.01 m of sparsely plagioclase-phyric basaltic trachyandesite recovered from Unit 17 consist of a brecciated top over a massive interior. The breccia is highly altered but only moderately fractured and disturbed, allowing the identification of most features. However, the alteration gives a deceiving view of the clasts. The centers of the clasts are dark gray (N3) but have 1- to 2-cm-thick dark reddish gray (10R 3/1) rims the same color as the matrix (Fig. F38). Between 70% and 80% of the breccia is composed of clasts >1 cm in size, <1-mm fines make up ~5%. The larger clasts are subangular to subrounded, and vesicularity ranges from 3% to 25%. Vesicle size is almost exclusively <1 mm, but one clast has a 1-cm-long elongated vesicle. The high degree of alteration hides the internal features of the clasts.
In Section
183-1139A-68R-7, from 140 cm downward, the lava is generally coherent, massive,
and aphyric. In fact, the lava is remarkable in its almost complete lack of
features. The massive lava contains 0.1
vol% vesicles that are 0.5
mm in diameter. Some 2- to 3-cm irregular vesicular pods appear in interval
183-1139A-69R-1, 105-150 cm, and a few 3- to 5-cm irregular voids appear further
down. There is no sign of increased vesicularity toward the base of the
recovered rocks from Unit 17.
The 8.77 m of moderately feldspar phyric trachyandesite recovered from Unit 18 contains a brecciated top, a massive interior, and a brecciated basal section. Unit 18 is also highly altered in places, masking many primary features (see "Alteration and Weathering"). Figure F39 plots vesicularity as a function of depth within Unit 18, but the identification of vesicles was commonly difficult because of the intense alteration. The breccia at the top is quite well preserved and contains many curious features (Fig. F40). The breccia generally has 60-70 vol% >1-cm clasts, and fines <1 mm make up ~10% of the breccia. However, the upper part has ~50% fine matrix and only 10% clasts >1 cm. This part of the breccia is matrix supported, and the clasts are rounded. Two other large, near-spherical clasts are present deeper in the breccia. These clasts have moderately vesicular margins. Inward of the vesicular margins is a sparsely vesicular concentric zone with large feldspar phenocrysts. The interiors of the clasts are again moderately vesicular. An interesting feature of the finer grained fraction of the breccia is the common presence of individual feldspar crystals in the matrix. The fractures within the clasts generally avoid the phenocrysts, and many small clasts are volumetrically dominated by a single phenocryst. Further into the breccia, alteration increases and the boundaries between clasts and matrix become indistinct in Section 183-1139A-70R-1, from 110 cm downward. There is no clear boundary between the brecciated and coherent lava. The massive interior generally has 1%-3% spherical vesicles 0.1-0.3 mm in diameter. Figure F39 shows the vesicular parts within the interior of the flow. Most of the visible color banding is caused by alteration.
The basal breccias of Unit 18 show distinctive features. In the lava, starting in Section 183-1139A-71R-3, at 47 cm, 1- to 2-cm rounded equant clasts appear. A 0.5- to 1-cm-thick zone of angular breccia with open void space is present in Section 183-1139A-71R-3, 58 cm (Fig. F41). Below this is a breccia containing dense lava with semicircular curved clasts and a matrix of red and very light green fine-grained material (Fig. F41). The lava below this breccia is made up of convoluted clasts with indistinct margins (Fig. F42). These clasts contain 0.5%-1% small spherical vesicles. Then, in interval 183-1139A-71R-3, 120-124 cm, another 0.5-cm-wide angular breccia zone with 1- to 4-mm fragments is present. Color banding is also prominent in several parts of this lower breccia. The very bottom of Unit 18 consists of a 15-cm-long lobate feature that has partially engulfed clasts of breccia underneath it. The feature contains 1-3 vol% rounded vesicles 0.1-1 mm in size. Elongated vesicles generally fan parallel to the margins of the lobate feature.
The 12.66 m of recovered moderately feldspar-phyric trachyte from Unit 19 is also highly altered with a breccia top over a more massive interior. The breccia on the top of Unit 19 shares many characteristics with the breccia at the top of Unit 18. Overall, clasts >1 cm make up 50-60 vol% of the breccia and <1-mm fines make up only 5%-10%. Clasts are rounded to subangular with no clear size sorting. Large clasts are distributed throughout most of the breccia but are not present at the base of the breccia. The uppermost clasts appear more rounded than those further inside the breccia. Lower in the breccia, the long axes of clasts are subhorizontal. In this same area, the edges of clasts become indistinct. Individual feldspar crystals are common in the fine-grained matrix.
The lava becomes coherent from interval 183-1139A-71R-5, 49-91 cm, downward. The interior of the flow is remarkably monotonous. Generally, cavities comprise ~1% of the interior, are 0.2-3 mm in diameter, have subangular shapes, and are filled with alteration minerals. Examination under the binocular microscope suggests that most, but not all, of these are altered phenocrysts. The only vesicle features we confidently identify are an irregular zone at interval 183-1139A-71R-6, 0-20 cm, and an inclined vesicle train in interval 183-1139A-72R-2, 31-70 cm. A 1.5 cm × 2 cm × >5 cm angular aphyric dark clast is present at interval 183-1139A-72R-1, 19-21 cm (Fig. F43). The last rocks recovered from Hole 1139A, in Section 183-1139A-73R-3 at 148 cm, were still in the massive interior of Unit 19.
In the pelagic and hemipelagic sediments of Units I through III, the volcanic component is very dilute. It is either incorporated into the pelagic sediments as primary pyroclastic air fall, and subsequently redistributed, or as suspended material from influxes of hemipelagic clay-sized material into the deep marine basin. Three discrete tephras were sampled from Unit II (intervals 183-1139A-11R-1, 107-109 cm; 38R-3, 17-19 cm; and 38R-4, 80-82 cm) (Table T9) and are believed to be bioturbated primary pyroclastic fall deposits.
The proposed environment of deposition for the sandy packstone (Unit IV) is a low-energy neritic setting, near a weathered volcanic source area (see "Lithostratigraphy"). This explains why a significant component of the clastic component of this sediment is volcanic lithics and clay. The grainstone (Unit 5) may have formed in a high-energy shallow marine carbonate shoal where the reworked terrigenous component, although not transported far, was overwhelmed by the influx of particulate carbonate material. Although the proportion of volcanic components in this unit is not great, the grain size and abundance of lithic fragments and crystal implies the sediment was deposited near an eroding volcanic terrain.
The pebbles and cobbles in Unit 1 are abraded and subrounded (Fig. F10A), and their surfaces are oxidized or weathered, implying that they are clasts and have been reworked, probably in a fluvial or littoral marine environment. Poor recovery in this interval (interval 183-1139A-49R-1, 109 cm, to 52R-1, 56 cm) does not allow us to attribute a thickness to a felsic lava, despite the recovery of several felsic lava clasts. Rounding and surficial weathering of the clasts provides evidence that the pebbles and cobbles form part of a pavement or conglomerate. Identification of this massive and flow-banded lava debris as beach pebbles and cobbles is consistent with the stratigraphic interpretation of the sedimentary sequence at Site 1139 (see "Lithostratigraphy"). These pebbles and cobbles may form part of a poorly consolidated conglomerate, from which the matrix has been lost during drilling. Several cobbles from this interval fell downhole during drilling and are present at the top of lower sections.
The bioclastic sandstone resembles the grainstone (Unit V) at the base of the sedimentary section and the environment of deposition is considered similar. Both units formed in a high-energy shallow marine environment (carbonate shoal?) where the reworked terrigenous component was overwhelmed by the influx of biogenic fragmental material. Although the proportion of crystals, glass, and volcanic lithic fragments is low, the presence of this volcaniclastic detritus indicates that the bioclastic sandstone was deposited near an eroding volcanic region.
Although little of this unit was recovered (47 cm), the largely clast-supported breccia has angular to subangular, pale green, altered, flow-banded felsic glassy clasts in a fine-grained matrix (Fig. F11). The matrix contains fine-grained glassy material that may have been derived from the grinding of clasts against each other. The clasts are not welded, indicating that they were not hot at the time of breccia formation. Many clasts are now banded perlites (Fig. F13), indicating the glassy nature of the clasts before devitrification. Some clasts appear pumiceous in core, but pumice was not observed in thin section. Banded perlite may impart a pseudopumiceous texture to the clasts. This breccia may either be the upper part of a autobreccia associated with a felsic lava flow, or a pyroclastic flow-top breccia.
The exquisitely developed spheroidal perlitic fractures (Figs. F13, F14) in this interval indicate that this rock was dense glass before hydration and alteration. However, abundant lithic fragments in this interval are unusual for a lava flow, so we interpret this unit as the vitric core of a densely welded pyroclastic flow deposit (vitrophyre). The gradational transition from the glassy zone into a silicic breccia with lithic clasts at the base of the interval (Figs. F17, F18), and possibly also toward the top, is consistent with this interpretation. A welded pyroclastic flow is strong evidence for primary deposition in a subaerial environment.
In the lower part of this interval (interval 183-1139A-54R-1, 56-108 cm), some clastic textures are preserved in the highly altered rocks. Pale green, wispy pumice-like clasts are present in some places (interval 183-1139A-54R-1, 56-62 cm) and bedded angular granules in others (interval 183-1139A-54R-1, 67-79.5 cm). In the upper part of the sequence, the rock is intensely sheared to resinous and powdery, green, clay-rich materials. The sheared textures and intense alteration suggests that this is a fault zone (see "Alteration and Weathering").
The drape, flattening, and welding textures of Unit 2 (Fig. F19) imply that it was formed by the agglutination of hot vesicular, glassy particles. The rocks have been identified as rhyolite on the basis of the sanidine (~10%) and quartz (~1%) phenocrysts. The welded rock is dark red (oxidized) and highly altered; glass shards are absent in thin section (Fig. F20). However, the rock has a flow-banded, draping, and flattening texture, with trains of small (<1 mm) irregular vesicles aligned subparallel to the flattening fabric. This material could have formed either as a primary spatter deposit or as a partially welded breccia associated with a hot pyroclastic flow deposit. We did not observe primary spatter fragments and bombs in this interval, suggesting that partial welding of a felsic pyroclastic flow is the more likely mechanism of formation. The dark red (oxidation) color of this interval may be a result of eruption and weathering in a subaerial environment.
The contact between the sanidine-phyric, partially welded breccia (Unit 2) and crystal vitric tuff-breccia (Unit 3) is defined by a color change from dark red to green. However, recovery of the contact is poor, so the nature of the transition between the two units is not clear.
The crystal vitric tuff-breccia is green, highly altered, and appears very different from the units that enclose it. However, the crystal content in Unit 3 (~15% sanidine and ~4% quartz) is similar to that in the dark red, vesicular, welded felsic rocks above (Unit 2) and below (Unit 4) (see "Igneous Petrology"). The perlite identified in thin section is diagnostic of originally coherent felsic glass. The preservation of domains of more lithified material with diffuse banding (Fig. F21), and lithic fragments distributed throughout the section, suggests that this probably was not a lava. We infer that this is a pyroclastic flow deposit, but the felsic component has been altered, and the only preserved clasts are larger and indurated. This observation, together with ubiquitous spheroidally fractured (perlitic) glass (Fig. F22), implies that this unit was a densely welded glassy interval. This is supported by faint banding in more massive parts of the core (interval 183-1139A-56R-3, 130-150 cm) that are not associated with enclosed clasts. The unit was probably a welded zone in the core of a pyroclastic flow deposit.
Unit 3 looks different from the glassy perlite identified in Subunit 1D, largely because the former is much more intensely altered. However, at the upper contact between Units 2 and 3, where color changes (dark red to green), there is evidence of shearing, indicating faulting. Perhaps this contact was a fluid percolation pathway leading to alteration of the glassy material.
We define the contact between the crystal-vitric tuff breccia and alkali feldspar-bearing, partially welded breccia by a bold color change from green (Unit 3) to dark red (Unit 4). However, recovery over the contact is poor so the nature of the transition between the two units is not clear and may be gradational. We cannot determine if there was a break in time between their deposition.
Poor recovery of Unit 4 and repeated brecciation make it a difficult section to interpret. However, the most common type of brecciation (Section 183-1139A-60R-1) with cataclastic textures could be a tectonic feature. The jigsaw fit across some of the clasts in this breccia suggests fragmentation in place with minimal transport. The other two well-cemented matrix-free breccias are likely to be welded. The uppermost breccia looks pumiceous and may be a slightly vesicle-poor pumice. Subvertically aligned elongated vesicles in the more massive rock may follow sutures where large (<10 cm), hot clasts were pressed together (Fig. F23). However, the lack of rheomorphic textures (e.g., laterally continuous flow banding) and the random orientation of feldspar phenocrysts argues against significant shear in the recovered parts of Unit 4. Both large (up to 10 cm) clasts and angular vesicles point to a very viscous lava, supporting the suggestion from the phenocryst assemblage that this is a felsic pyroclastic deposit.
The Unit 4 to Unit 5 boundary lies within a brecciated zone showing a change from a matrix-supported interval at the base of Unit 4 (Fig. F24) to a hydraulically fractured and altered breccia at the top of Unit 5 (Fig. F25). The precise contact is not clear.
The upper breccia in Unit 5 (Section 183-1139A-61R-1 [Pieces 1 and 2]) (Fig. F25) is, at least partly, brecciated as a result of crosscutting tectonic fracturing and veining during alteration. The random orientation of the large feldspar phenocrysts suggests that the recovered lavas were either disrupted and mixed by brecciation or were not subjected to high shear during crystallization. In thin section, the massive zone has a lava-like texture (see "Igneous Petrology") with microcrystallites in the mesostasis. The fractured zones within the massive part of the unit may be related to cataclasis and subsequent alteration (see "Alteration and Weathering"). The recognition of a massive central zone with breccias above and below, and the thin section evidence that the massive interior have lava-like textures, supports the identification of this unit as a lava flow.
The change from a feldspar-phyric trachyte to an aphyric trachybasalt is a clear boundary. However, no contact was recovered, making it impossible to interpret the relationship between the two units.
The recovered rocks from Unit 6 (Fig. F26) indicate a lava with a dense interior and a more vesicular breccia top. However, it is not clear that the breccia on the top of Unit 6 is an autobreccia formed during emplacement; it could be an eroded and reworked vesicular flow top. The breccias within Unit 6 are probably tectonic in origin, but we cannot rule out the possibility that one or more are flow-top or basal breccias. If any are autobreccias, Unit 6 represents more than one package of lava.
The morphology of the few well-preserved clasts within the breccia at the top of Unit 7 indicates that it formed during the emplacement of the flow. Thus, it must be part of a flow-top or basal breccia, indicating proximity to a flow boundary. However, the location of the unit boundary at the top of Core 183-1139A-64R is arbitrary. The actual contact is probably in the unrecovered portion of Core 183-1139A-63R, or somewhere within the heavily altered and fragmented rocks in Unit 6. Several points suggest that the breccia at the top of Unit 7 is actually a basal breccia to a flow that is mostly in Unit 6. Given the equivocal nature of the contact between Units 6 and 7, we do not attempt to interpret relationships between the two units.
The fluidal shapes and the possible entrainment of sediments into some trachybasalt clasts argues for brecciation or disruption during emplacement. The fine-grained margins suggest chilling where the fluidal clasts were in contact with colder fragments of lava. The mixing with sediments and the horizontal (flattened) shapes of the fluidal clasts suggest a basal breccia. Alternatively, it could be a mix of spatter and lithics or even a lava-rich peperite. If these clasts are spatter, they have unusually low vesicularity. This might suggest an explosive interaction between somewhat degassed lava and near-surface water. The simplest interpretation is that this is a welded basal breccia. The shapes of the vesicles within the coherent part of Unit 7 indicates significant shear. We interpret the rectangular vesicular domain to be an entrained clast, suggesting entrainment of a disrupted upper crust. The small xenolith could be from the magma chamber or incorporated during flow, especially if the flow started as a mix of spatter and lithics. It is not clear if the breccia at the base of Unit 7 is related to emplacement or postemplacement processes. Given the equivocal nature of the breccia at the top of Unit 7 and the lack of information about its base, we cannot draw any definitive conclusions about the type of flow it contains.
We placed the boundary between Units 7 and 8 at the point where the color of both the clasts and the matrix change distinctly within the highly fragmented breccia. It is not clear that this location represents any change in lava types, but it is a visually distinct boundary in the breccia that lies between two different flows.
The highly altered, fragmented, and disturbed breccia at the top of Unit 8 is difficult to interpret. The relatively low vesicularity of the identifiable clasts as compared to the top of the better recovered and preserved lava is particularly puzzling. We interpret the zone of variable vesicularity as a zone of welded clasts. The shapes, sizes, and high vesicularity of the clasts suggest primary spatter. Because the zone of variable vesicularity (i.e., welded clasts) grades into the coherent lava underneath, it is unlikely that this is a welded basal breccia of the overlying unit. The horizontally elongated, but well-rounded shapes of the vesicles in the coherent part of the lava suggest shear during crystallization of a relatively low viscosity lava. The increase in vesicle size with depth is consistent with increasing coalescence with time. We interpret the increased vesicularity toward the base of the unit to be the basal vesicular zone. The centimeter-thick zone that has altered to a dark gray may represent the basal chill.
Although the clasts in the upper part of Unit 8 may be spatter, proximity to the vent is not required. The welded spatter could easily have been transported tens of kilometers as a raft on an open lava channel. Also, the degree of coalescing implied by the very large vesicles and their sheared shapes suggest that the lava was transported before being frozen. It is not possible to confidently determine the lava flow type from these observations.
The breccia at the top of Unit 9 indicates a boundary between two separate flows. The continuous groundmass size and the complete cementation of the clasts without any matrix suggests welding. Folded shapes suggest that the clasts were hot and plastic at the time of formation. The high density of very small vesicles at the base of breccia suggests a chill margin, perhaps indicating that it is more likely to be the base of Unit 8 rather than the top of Unit 9. However, given that only one isolated 10-cm piece of breccia was recovered, we cannot interpret relationships between Units 8 and 9 and the breccia unequivocally.
Interpreting the vesicle distribution within Unit 9 (Fig. F28) without recovery of a flow top or flow base is difficult. The relatively low vesicularity and subvertical elongation of the vesicles in the uppermost coherent lava is difficult to reconcile with a flow top; a significant amount of material may be missing from the top of the flow. The increased vesicularity and a switch to angular shapes would be easiest to explain as a zone of high shear, but this would have disrupted the crust above. An alternative is an entrained aa-like clast, but there is no evidence for clast boundaries, and this zone is much wider than typical entrained clasts. The zone of elongated but rounded vesicles below suggests a reduction in shear and a more fluid lava, but steeply inclined vesicle trains suggest significant nearby topography within the flow. The zone of increasing vesicle size with constant vesicularity (by volume) suggests increasing coalescence of vesicles.
The top of Unit 10 is an altered and sheared breccia with some clasts similar to the lava in Unit 9. The contact between the lava flows of Units 9 and 10 could be anywhere within this breccia, and we cannot interpret the relationships between these flows.
The breccia at the top of Unit 10 is difficult to interpret. The fluidal shapes and higher vesicularity of some clasts could suggest spatter or a partially welded basal breccia, but given the degree of alteration, this material could have originally been a coherent flow. Within the better recovered lava, we interpret the zones with relatively low vesicularity, but high vesicle number density as altered, welded chill zones (Fig. F29). In general, the lobes separated by these chill zones have a vesicular upper crust, dense interior, and vesicular lower crust, as is typical of inflated pahoehoe flows. The boundaries between the upper vesicular crust and massive interior are unusually sharp for lobes this small. Smaller lobes tend to cool too quickly for the bubbles to segregate effectively from the stagnant interior of the flow. This suggests that either the lava was relatively fluid or that the lobe remained hot unusually long. This could be readily achieved if the flows were rapidly emplaced on top of each other, producing a single thick "cooling unit." The apparently welded contacts between the lobes support this idea. The somewhat circular vesicular patches in Section 183-1139A-65R-2 may be a stack of small pahoehoe lobes welded together, or they could be the transition from a brecciated flow top to coherent lava. This might suggest that the uppermost lobe was emplaced too rapidly to form a coherent pahoehoe crust. The zone of poor recovery at the top of Core 183-1139A-65R-3 could be the result of vesicle sizes approaching the size of the diameter of the core (Fig. F29). The clast near the base of the second lobe is probably a folded piece of the lower margin, possibly bent by local 5- to 10-cm-scale topography. This kind of folding again suggests relatively rapid emplacement of pahoehoe lava.
It is interesting that some features common in inflated pahoehoe flows are missing from Unit 10. In particular, there is no evidence for effective segregation of late stage liquids into vesicle cylinders and horizontal vesicle sheets. Instead, only wispy mesostasis blebs are present in the largest of these lobes. We speculate that these lobes did not have sufficient volume to segregate large bodies of late-stage volatile-rich melt.
The base of Unit 10 is a clear boundary (Section 183-1139-65R-3, 146 cm) with a smooth pahoehoe surface partially engulfing breccia clasts. The breccia clasts are subrounded, suggesting sufficient time for secondary processes to smooth the clasts.
The breccia on the top of Unit 11 is too poorly recovered and too reworked/weathered/altered to identify the style of brecciation. However, the mixing of vesicular and denser lavas in the upper part of the coherent lava is diagnostic of entrainment of breccia clasts. This means that the breccia at the top of Unit 11 is an autobreccia and is not simply the breakdown product of a coherent lava. The morphology of clasts in the well-preserved basal breccia definitively identify Unit 11 as an aa flow. The vesicles in the upper part of the flow have the irregular, angular shapes typical of aa flows, but the interior includes more rounded vesicle shapes. This suggests that the interior of the flow was still somewhat fluid when the flow stopped. Wispy mesostasis blebs without larger segregation features are also consistent with most aa flows. The entrained clast with the large irregular vesicle at its top is very unusual (Fig. F33). We speculatively suggest that it may represent air or water (steam) incorporated into the lava when the clast was entrained into the flow. The vesicular domains in the lower part of the flow suggest that the basal breccia was also entrained into the interior of the flow.
Although a contact clearly exists between the massive interiors of Units 11 and 12, the basal breccia of Unit 11 and the flow top breccia on Unit 12 are too highly altered to accurately locate the boundary. The placement at a section break was convenient but arbitrary. Intense alteration precludes the determination of the precise contact between Units 11 and 12.
Unit 12 is a confusing lava that combines aspects of both aa and pahoehoe flows. The entrained clasts in the upper part of the coherent lava of Unit 12 require that the lava had a disrupted crust at the time of emplacement. The identifiable clasts in the breccia at the top of Unit 12 are angular but rather small to be true aa clasts. Also, the vesicles are not as angular as is typical in aa clinker. There is no straightforward explanation for the megavesicles and the zone of high vesicularity within the otherwise massive interior of the flow. Megavesicles imply coalescing of vesicles. How the vesicles were able to coalesce to form the megavesicles while coalescence was halted in the vesicular zone is unclear. The base of the flow poses additional puzzles. The dark groundmass is interpreted to be glassy margins that have altered to clay. These margins suggest a stack of 10- to 20-cm-thick pahoehoe lobes at the base of Unit 12. Relating these lobes to the brecciated flow top is difficult. Also, the contorted laminations in the sediments suggest that the lava plowed into the sediments and curled them over before baking them (Fig. F34). Retaining the original sedimentary laminations requires relatively gentle emplacement of the lava and suggests that the sediments were not very thick, probably <10 cm. Although the sediments may have been moist, the lack of disruption from steam generation implies minimal amounts of water or the emplacement may have been slow enough to gently boil off the water.
There clearly is a contact between the massive interiors of Units 12 and 13. However, given the extensive alteration and disturbance of the breccia between the two massive parts, the exact location of the boundary is equivocal. In this case, we assigned the last somewhat recognizable pieces of lava to Unit 12 and placed everything else at the top of Unit 13. Although these rocks do not allow us to directly interpret relationships between Units 12 and 13, sediments in the lower part of Unit 12 imply a time break between the emplacement of these two flows.
Unit 13 clearly had a brecciated flow top, but its features are not diagnostic of an aa flow. The first well-preserved piece of lava from Unit 13 has a vesicularity identical to the upper part of the coherent lava under the breccia, suggesting that an arm of the lava interior may have reached into the breccia. This arm contains more vesicles, especially larger ones, than dense arms intruding aa breccias commonly do. Clasts in the breccia now appear rounded, but the surrounding angular fragments suggest that this rounding likely results from in situ alteration rather than sedimentary reworking. The gradual decrease in vesicularity and vesicle size is consistent with a combination of larger vesicles rising and lithostatic pressure compressing the lower bubbles. The gradual increase in roundness of the vesicles with depth hints at a gradual stagnation of the flow. The basal section suggests shear, and a basal breccia could be expected below the base of Unit 13.
A contact must exist between the massive interiors of Units 13 and 14. However, the exact location of the boundary is arbitrary, and, for convenience, we selected the top of Core 183-1139A-67R as the beginning of Unit 14. It is possible that some of the breccia at the top of Unit 14 is actually a basal breccia for Unit 13. Given the state of the recovered rocks, we did not interpret the relationship between Units 13 and 14.
The relatively unaltered and undisturbed portion of the breccia in the upper part of Unit 14 is the first (partially) well-preserved flow-top breccia in Hole 1139A. However, we can also draw some important conclusions from the overlying altered part. First, an autolith is diagnostic of a breccia disrupted by melt during emplacement (Fig. F35). Second, the lack of angular vesicles (despite the wide range of observed vesicle morphologies) argues against a classic aa breccia. The elongated vesicles along the margins of the larger clasts are more typical of pahoehoe lobes or spatter. However, the moderate vesicularity and large size of the clasts argues against the clasts resulting from spatter.
The most striking feature of the well-preserved breccia is the gradational increase in the fine-grained matrix with depth and its gradual change to a coherent lava. Two explanations are (1) the lava contained a large amount of fines that became sintered into a coherent lava deep in the breccia or (2) the coherent lava disaggregated as it cooled higher in the breccia. The observations weakly favor the first possibility. Highly rounded equant shapes of the larger clasts suggest much mechanical abrasion that would produce many fines. Also, where we have observed disaggregating dense lava in Hole 1138A and Hawaii (L. Keszthelyi, unpubl. data), the transition occurs over a distance of a few centimeters, not the several tens of centimeters as observed here. In either case, the indistinct appearance of the clast margins deep in the breccia indicate gradual assimilation into the coherent lava.
The generally low vesicularity of the massive interior of Unit 14 suggests efficient degassing during emplacement of the flow. The lowermost interior with mesostasis blebs appears to record how the few small vesicles formed in the massive interior. These vesicles may have been escaping from the late-stage liquids represented by the mesostasis blebs. Their small size may have kept them from rising rapidly enough to escape through the disrupted upper crust before the flow stopped and solidified.
The basal part of Unit 14 records a process similar to that in the upper breccia. The gradual decrease in dense coherent lava and increase in the distinctness of the vesicular clasts with depth suggests that a basal breccia was being remelted and assimilated along the base of the flow interior. The lava in the highly fractured, altered, and disturbed breccia below this level is not clearly associated with Unit 14. In particular, the clasts with angular vesicles have no counterparts in the flow-top breccia. Also, mesostasis blebs are very rare in smaller lobes and are usually confined to the massive interiors of flows.
For the reasons noted above, some of the breccia under the coherent part of Unit 14 may actually belong to Unit 15. Given the equivocal nature of this breccia, we make no attempt to interpret the relationship between Units 14 and 15.
The recovered rocks from Unit 15 show the features of the middle and lower parts of a lava flow. However, all we can infer about the flow top is that the basal brecciated zone assigned to Unit 14 implies a flow top breccia. The change from irregular to rounded vesicles at the top of the recognizable rocks in Unit 15 might be the transition from the upper coherent vesicular crust to the massive interior, suggesting that we have the entire interior of the flow. The vesicular zone within the massive interior exhibits features atypical of horizontal vesicular zones in inflated pahoehoe flows (see "Physical Volcanology" in the "Explanatory Notes" chapter). In particular, the changes in vesicle shapes but relatively constant vesicularity (in volume percent) and vesicle size are atypical. The gradation into mesostasis blebs at the base of the vesicular zone suggests that some of these vesicles may have formed during crystallization of the flow. The gradation from patchy vesicular zones to a coherent vesicular lava at the base of the flow suggests a welded basal breccia that has been disrupted by and incorporated into the floor interior.
The nature of the base of Unit 15 suggests a basal breccia. However, for convenience we include all the breccia below the coherent part of Unit 15 in Unit 16. Thus, the location of the Unit 15/Unit 16 boundary probably does not reflect the contact between the two flows. However, no features in the breccia indicate exactly where the contact is. Given the nature of this boundary, it is clear that no significant weathering horizon or sedimentary interbed formed between the two flows. This suggests that the two units were emplaced closely in time.
The better preserved parts of the breccia at the top of Unit 16 strongly indicate that it formed during emplacement of the flow. In particular, the unsorted mix of lithologies and angular to subangular shapes are not consistent with extensive reworking of the breccia. The clasts do not appear to include classic aa clinker but do include small partially fragmented pahoehoe lobes. The large dense clast (interval 183-1139A-68R-5, 100-116 cm) is likely to be an arm of the massive interior that has penetrated into the flow-top breccia. The banding in the arm is more reminiscent of felsic flow banding than that in basaltic flows. This suggests that the arm was more viscous than expected for basalt. The relatively poorly preserved massive interior appears similar to the interiors of other breccia-topped units though entrained clasts are notably absent. The slight increase in vesicularity at the base of the massive lava might be the start of a basal vesicular zone.
The contact between Units 16 and 17 is probably in the breccia, but its exact location is equivocal. We have chosen to place all the breccia at the top of Unit 17, even though some of this material may be a basal breccia for Unit 16. However, the breccia immediately below the coherent interior of Unit 16 is highly altered, and no contact between a flow-top and flow-base breccia can be discerned. Given the altered nature of the breccia, we cannot determine exact contact relationships between the two units.
The high degree of alteration of the breccia at the top of Unit 17 makes detailed interpretation difficult. In fact, there is no strong evidence that this breccia was formed during emplacement, but the clasts are generally similar to the autobreccias on the other flows. There is no apparent reason why the interior of Unit 17 is more massive and featureless than the interiors of the overlying units. The large irregular voids are as likely to have formed by alteration processes as during the original emplacement of the flow.
The change from the sparsely plagioclase-phyric interior of Unit 17 to the moderately sanidine phyric top of Unit 18 is clear, but the contact was not recovered. The matrix supported conglomerate at the top of the Unit 18 breccia could be the result of reworking of a flow-top breccia. This tentatively suggests a significant time gap between the emplacement of Units 17 and 18.
The breccia on the top of Unit 18 is difficult to interpret. Although there is no definitive evidence that Unit 18 is a lava flow, the low vesicularity of the clasts in the breccia argues against a pyroclastic origin. Also, the largest clasts appear to be 10- to 20-cm-diameter lobes. Shear during flow through a lobe can mechanically push large phenocrysts to the lobe margin, possibly explaining the crystal-rich annuli. This suggests that the breccia top to Unit 18 might be similar to some of the transitional flows seen at earlier sites (see "Physical Volcanology" in the "Site 1137" chapter and "Physical Volcanology" in the "Site 1138" chapter). Brecciation has also apparently separated the large feldspar phenocrysts from the glassy matrix of some clasts. Resedimentation of the fines from breccias like those on top of Unit 18 could produce the feldspar-rich sands seen at various locations in Hole 1139A. The blurring of clast margins at the base of the breccia could be completely the result of alteration but also suggests some welding deep in the breccia.
The pattern of vesicularity in the interior of Unit 18 does not lend itself to a straightforward interpretation, especially because the identification of vesicles was commonly very difficult. However, the basal breccia shows a complex history and a wide range of rheologic behavior. The lobe at the base of Unit 18 suggests that the leading edge of this trachyte flow might not have been very different from a pahoehoe flow. The welded, contorted clasts above this lobe indicate that the lava that rode over the initial lobe was being disrupted. It also suggests that the breccia being shed from the front of this second wave was covered quickly enough for the clasts to be unable to freeze solid. The breccia with hemispherical clasts could be a trachytic version of slab pahoehoe top and would indicate that the lower ~1 m of Unit 18 is physically separate from the large overlying lobe. Finally, the narrow zones of angular breccia show evidence for shear (presumably continued flow of the lava) after the flow had cooled to the point where it could undergo brittle failure.
The breccia at the top of Unit 19 has the same character as the breccia at the top of Unit 18, so we interpret it to be the flow top breccia for Unit 19. The increased rounding of the uppermost clasts again suggests of some significant time break between the emplacement of Units 18 and 19.
The interpretation of the breccia on Unit 19 poses similar problems as for Unit 18. However, in this case, lobe-like clasts are absent and the horizontal elongation of clasts suggest some flattening of the breccia. This makes it impossible to rule out a pyroclastic origin for this flow-top breccia. The massive, highly altered interior of the flow does not provide much to interpret. The curious lithic fragment (interval 183-1139A-72R-1, 19-21 cm) in the lava (Fig. F43) is very difficult to explain if it was not thrown onto the top of the flow by explosive activity and then entrained into the flow as it assimilated its brecciated crust. Xenoliths of such dense rock are expected to sink to the bottom of the flow rather than be suspended in the upper part of the interior of the flow.
Units 2, 3, and 4 might plausibly represent different welding zones within a single thick pyroclastic flow deposit. This interpretation is consistent with the similar phenocryst assemblage, abundance, and size in all three units (see "Igneous Petrology"). In addition, Units 2 and 4 have very similar physical characteristics (see "Physical Properties" and "Downhole Measurements"). The lower K2O content of Unit 3 relative to Unit 4 is probably caused by differences in the type of alteration, as Unit 3 is pervasively altered to green clay, whereas Units 2 and 4 are red and oxidized and probably have been silicified (see "Igneous Petrology" and "Alteration and Weathering"). Welded pyroclastic flow deposits only rarely form in submarine environments, and welding in such deposits is not sufficiently intense to form vitrophyre. Thus, the pyroclastic flow deposit in Units 2 through 4 was very likely emplaced subaerially. The oxidation of Units 2 and 4 further indicates weathering in a subaerial environment for a significant period of time following emplacement. However, given the poor recovery and intense alteration, we cannot reach definitive conclusions.
Given the highly altered and disturbed states of most contacts within the lava pile, it is difficult to make any confident generalizations about their style of emplacement and the environment they were emplaced into. However, there is no evidence for interaction with significant amounts of water. Also, the basaltic flows are probably <10-m-thick, making it unlikely that they traveled >100 km from their vents.
