The total depth penetrated at Site 1100 was 100.9 mbsf (Hole 1100D), and ~4 m of sediment was recovered (4.3% recovery) from one core (Core 178-1100C-1R, 0-5 mbsf). Recovered sediment consists of poorly consolidated massive diamict (see "Lithostratigraphy" in the "Explanatory Notes" chapter), with a diatom-bearing silty clay matrix (Fig. F5). Site 1100 lies at the edge of an area of hummocky, ice-scoured topography underlain by an acoustically transparent facies with indistinct layering and numerous point-source diffractions, interpreted by Vanneste and Larter (1995) as till.

The character of the diamict recovered at this site is consistent with iceberg plowing of underlying till (e.g., Woodsworth-Lynas and Dowdeswell, 1993; Pudsey et al., 1994). Rare, open-shelf marine microfauna are reworked and abraded, consistent with faunal data collected from tills at Site 1097 (see "Biostratigraphy"). No age inference can be made from any of the material examined because of the poor nature of its preservation.

The core recovered at Site 1100 is compatible with currently available sedimentological data from the Antarctic Peninsula shelf, which are limited. Short (maximum length ~4 m) piston cores were taken from the shelf area by Pudsey et al. (1994) and from the area to the south by Pope and Anderson (1992). Several cores were collected near Site 1097. Sediments from these cores were described as "terrigenous gravelly muds" and interpreted as "compound glacial marine sediments" by Pope and Anderson (1992), meaning marine muds with ice-rafted debris. Such facies contain a low-diversity foraminiferal fauna with reworked sponge spicules and diatoms. Pudsey et al. (1994) also described a "structureless diamicton" transitional to laminated marine muds from a piston core site nearby. All these facies are regarded as the tops of tills reworked postdepositionally by iceberg turbation.

Poor core recovery at Site 1103 allows a partial picture of sedimentary history to be established (Fig. F6). The total depth of the hole was 362.7 mbsf. Recovery from the upper 247 m was 2.3% but improved in the lower 115 m, where 38.7 m of sediment was recovered (34% recovery). This sediment consists of diamictites, poorly sorted sandstones, and mudstones that are interpreted as sediment gravity flows. In the absence of a continuous sedimentary record, stratigraphic subdivision is not possible. Nonetheless, several sedimentary facies can be identified. These record deposition of debris flows and poorly sorted turbidites on an active, glacially influenced slope. Biofacies data are insufficient to constrain age, water depths, or position relative to the shelf break.

Three lithofacies can be identified at Site 1103. These are massive and chaotically stratified diamictite (50% of recovered core), sandstone (25% of recovered core), and mudstone (25% of recovered interval) (Fig. F7).

Diamictites are poorly sorted, lithified mixtures of clasts, sand, and mud (see "Lithostratigraphy" in the Explanatory Notes" chapter). At Site 1103, diamictites have a distinctive gray, asphalt-like appearance (Fig. F8A, F8B). Single beds of massive, unstructured, and matrix-supported diamictite (Facies Dmm) range from 20 cm to 4.2 m in thickness (e.g., Sections 178-1103A-28R-2 through 28R-5; Fig. F7A, F7B, F7C). Beds are separated by thin (<30 cm) intervals of chaotically stratified facies (Facies Dms) showing highly deformed "flow-banding" (Visser, 1993) defined by "stringers" and "streaks" of mudstone (e.g., Sections 178-1103A-28R-1 and 28R-2; Figs. F7B, F8C, F8D). Diamictites vary from clast rich (>20% gravel) to clast poor (10%-20%), with the latter showing a smaller overall clast size (<2 cm). Gradations between poorly sorted muddy siltstone with dispersed clasts (Facies Fmd) and diamictite (Dmm) occur in several cores (Cores 178-1103A-33R and 37R; Fig. F7E, F7I). The largest clast size observed is 10 cm, and clast shape varies from angular to rounded. Measurements of clast long-axis dip angle show a wide variation from flat-lying to steeply dipping clasts (Fig. F9); preferred orientation of clasts and matrix is apparent in some thin (<20 cm) intervals that show a "flow fabric." Clasts are supported by a poorly sorted muddy sand matrix (ranging from 60% sand, 30% silt, and 10% clay to 30% sand, 50% silt, and 20% clay). The sand fraction is compositionally immature, with >40% lithic fragments.

A distinctive feature of diamictites at Site 1103 is the presence of large amounts of reworked marine mud in the form of discrete clasts and deformed matrix. Small mudstone clasts (<2 cm diameter) (variably termed "intraclasts," "rip-up clasts," "silt clasts," or "chips" in the literature [Pickering et al., 1989; Eyles et al., 1993]) are a ubiquitous component of diamictite facies (Fig. F8C) and in some intervals form as much as 20% of the total clast population, resembling the "clast breccias" of Pickering et al. (1989). In addition, a common component of these diamictites are irregular patches and stringers of dark mudstone. Examination of thin sections of chaotically stratified diamictite suggests that as much as 20% of the diamictite matrix consists of reworked mudstone. Mudstone clasts contain reworked Miocene marine biota and Cretaceous radiolarians (see "Biostratigraphy"). Shell fragments are common and include a complete barnacle valve in Section 178-1103A-28R at 120 cm.

Beds of chaotically stratified and massive muddy sandstone (Sd and Sm, respectively; Figs. F7, F10A) are from 2 cm to 3 m thick (Cores 178-1103A-37R and 38R; Fig. F7I, F7J). Sandstones are very poorly sorted (65% sand, 25% silt, and 10% clay) and gray in color. Bed bases are sharp and erosional; bed tops were not recovered. Chaotically stratified sandstones show extensive soft-sediment deformation and incomplete mixing of different grain sizes (Fig. F10A). Thin (up to 2 cm) graded intervals of conglomerate and mud clasts are present at bed bases (Fig. F7J). Section 178-1103A-32R-1 (Fig. F7D) consists of a thin (38 cm) interval of parallel-laminated sandstone, with successive laminations fining upward and showing small-scale "convolute" soft-sediment deformation. Thin beds of steeply dipping sandstone occur in Core 178-1103A-33R interbedded with mudstone and diamict (Figs. F7E, F10B). Floating granules, gravel, and mud clasts are present throughout the sandstone facies, and shell fragments are present in all cores.

Fine-grained sediments at Site 1103 are very poorly sorted silty mudstones. The dominant facies is massive (Facies Fm; Fig. F11A), with a subordinate weakly laminated facies (Facies Fl; Fig. F11B, F11C). Laminations have no systematic structure but are defined by thin (<1 mm) streaks and sheared-out "blebs" of siltstone and mudstone. Mudstones show transitions from massive to laminated facies with no systematic trends upcore (Core 178-1103A-31R; Fig. F7D). Mudstones are very poorly sorted (5%-20% sand, 35%-70% silt, and 10%-60% clay), with the same gray color as diamictites and sandstones. "Floating" clasts are present in massive and laminated mudstones (Facies Fmd and Fld respectively; Fig. F7). Bioturbation is absent. Deformation structures, such as small-scale normal faults (e.g., intervals 178-1103A-34R-1, 127-130 cm; 35R-1, 45-50 cm, and 36R-3, 20-30 cm; Fig. F7) and dewatering structures (Fig. F7), are present throughout. Section 178-1103A-37R-3 shows pillow structures and associated downward-penetrating dikes below a thin (4 cm) bed of medium-grained massive sandstone (Fig. F7I). Section 178-1103A-31R-1 is composed of interbedded laminated and massive mudstone, with dispersed clasts, and diamictite (Fig. F7D). In other cores (Core 178-1103A-33R; Fig. F7E), there are gradations among diamictite facies, mudstone with dispersed clasts, and mudstone lacking any clasts.

Massive diamict(ite) is not diagnostic of any one depositional environment and can be deposited in a wide range of settings. Interpretation relies on examination of other related facies to provide a context for deposition. Diamictites at Site 1103 are associated with massive sandstones and siltstones. The former can be identified as "disorganized muddy sands" of Pickering et al. (1989), deposited either by very high density turbidity currents or very fluid sand-mud debris flows. A muddy matrix creates sufficient buoyancy to enable clasts to be freighted within the flow, and grading is absent, except for thin "coarse-tail" graded intervals at bed bases (e.g., Fig. F7J). Chaotically bedded and massive, poorly sorted sandstones record very rapid deposition and rapid "freezing" of concentrated dispersions transported by sediment gravity flows. Horizontally laminated sandstones in Core 178-1103A-32R represent the B division of turbidites (Pickering et al., 1989).

Massive and irregularly laminated siltstones also record rapid deposition from dense turbidity currents or muddy debris flows; lamination is probably generated when the flow progressively "freezes" from the bed base to the top during deposition. Widespread soft-sediment deformation structures attest to rapid deposition and local fluidization generated by rapid dewatering. Additional deformation, such as faults, may result from downslope creep.

Stratigraphic successions of diamictites interbedded with poorly sorted massive sandstones and siltstones are classically associated with downslope resedimentation of poorly sorted debris by sediment gravity flows. Beds of massive diamictite at Site 1103 (Facies Dmm; as much as 4.2 m thick) that show variation in clast content and contain large amounts of reworked sediment such as silt clasts and mudstone are, consequently, identified as debris flows (debrites). Successive flows are probably recorded by the interbedding of massive facies with thin zones (<30 cm) of crudely stratified diamict with mud stringers (Facies Dms; Fig. F7). Muds were probably deposited on bed tops between flow events and were subsequently reworked and incorporated in later flows and deposited as chaotically stratified diamict facies (Dms) that show flow banding (Visser, 1993). The absence of bioturbation and the lack of any in situ marine biota from mudstones suggest short recurrence intervals between flows. Relatively coarse-grained turbidites, such as those recovered at Site 1103, are deposited rapidly; high overall sedimentation rates are indicated by common water-escape structures and soft-sediment deformation.

Sediments recovered at Site 1103 record an active slope close to a source of poorly sorted glacial debris, such as till. In common with other widely reported debrite/turbidite facies associations (e.g., Walker, 1992), diamictite, sandstone, and mudstone probably represent an evolutionary sequence recording progressively better sorting downslope and the loss of the coarse fraction (clasts) from debris flows in response to increasing water content and nonlinear processes. Consequently, these facies are intimately interbedded and show upcore transitions from clast-rich to clast-poor diamictite and from diamictite to pebbly mudstone to mudstone (e.g., Core 178-1103A-33R; Fig. F7E). Protracted downslope transport results in rapid expulsion of mud and the generation of well-sorted graded conglomerates, sandstones, and siltstone turbidites (e.g., Pickering et al., 1989; Walker, 1992; Eyles et al., 1993). Such facies are limited at Site 1103 to thin (<10 cm) sandstone beds in Cores 178-1103A-32R, 33R, and 37R (Fig. F7D, F7E, F7I). Because of the very poor overall sorting in turbidites at this site, an upper slope setting close to an area of active slumping is suggested. Biofacies data are limited and suggest an outer shelf setting (see "Biostratigraphy"). Unfortunately, age-specific material was not recovered. The abundance of reworked mudstone and marine biota is similar to the characteristics noted for deformation tills recovered at Site 1097 (see "Lithostratigraphy"  in the "Site 1097" chapter) and indicates a till source for debrites at Site 1103.

The sedimentary record at Site 1103 is consistent either with an upper continental slope setting or with deposition on the shelf, such as within an overdeepened trough or along an ice margin. Glaciation of continental shelves results in large subglacial fluxes of sediment to the outer shelf edge and the remobilization of muddy debris and existing shelf edge sediments downslope as slumps, debris flows, and turbidity currents. Such facies have been described from the modern Antarctic continental shelf and slope by Anderson et al. (1979, 1980, 1984). Detailed facies descriptions of debrite and turbidite associations from other settings have been presented by several workers (e.g., Visser, 1983; Hill, 1984; Miall, 1985; Young and Gostin, 1991; Eyles, 1990, 1993; Eyles et al., 1993; Gipp, 1993). It is possible that the debrite/turbidite facies association recovered at Site 1103 provides insight into sedimentary processes operating on the present continental slope and the probable composition of slope "foresets" underlying the present continental slope. Alternatively, such facies may be representative of the "till deltas" argued to be forming at the grounding line of Ice Stream B of the Ross Ice Shelf (Alley et al., 1989). The existence of similar till deltas on the continental shelf of the Antarctic Peninsula has been proposed by Vanneste and Larter (1995) on the basis of interpretation of deep-tow boomer profiles.

While waiting for the swell to subside at the continental shelf edge, JOIDES Resolution scientists undertook a survey of the seabed close to Site 1102, using the camera usually lowered around the drill string in order to effect reentry. The survey covered an area ~50 m × 50 m and took ~2 hr to complete. Two short sections of this survey (23:58:00 to 00:00:00 and 00:19:00 to 00:22:20) are presented as a QuickTime video (Movie M1). The view is of the seabed, illuminated by camera lighting; the drill bit (~28 cm diameter) and lower BHA remain a constant distance from the camera but move with respect to the seabed as a result of slow, regular ship heave and horizontal ship motion. The heave compensator was turned off during the survey; ship heave was estimated (from the acoustic signal of a pinger attached to the camera) as 3 m average, 5 m maximum.

The camera survey showed a carapace of coarse, subrounded-to-angular blocks as much as a meter across over much of the seabed. We saw sharp transitions from patches of mainly fine-grained seabed sediment, a few meters across, to larger areas of blocky seafloor (see, for example, 23:59:40). We could not establish shape or orientation of areas of different texture, or any sedimentary structures.

Having failed to spud in at Holes 1102A and 1102B, we used the survey to choose a better place (showing finer grained sediment at the seabed) to spud in at Hole 1102C. Neither of the succeeding holes was successful, however, and we conclude that a layer of coarse, angular blocks also underlies the areas of fine-grained seabed, the total thickness probably exceeding 10 m.

Such a carapace has not been seen or sampled elsewhere on the continental shelf. We speculate that the blocks were transported to the shelf edge within subglacial till at times when the ice sheet was grounded to the shelf edge, with the likely addition of postglacial ice-rafted debris. Subsequent proglacial processes (plowing by iceberg keels, with the suspension of fine-grained sediment and its transport across the nearby shelf edge by tides and currents) have left only a coarse lag deposit on the outermost shelf. We are uncertain if such a coarse, shelf-edge facies is preserved in situ within the geologic record of the glacial prograded wedge or if it is moved onto the upper slope during the next grounding line advance.