A slippery slope for Cryogenian diamictites?

The Death Valley region has previously been claimed to preserve the sedimentary records of both the Sturtian and Marinoan snowball Earth events within the Kingston Peak Formation, which outcrops in a number of disconnected mountain ranges. In this context, new sedimentary logs are presented together with detailed clast textural analyses which allow diamictites of the Alexander Hills and the Saddle Peak Hills to be compared in detail for the first time, and to be contrasted with rocks of well‐established glaciogenic origin from the Kingston Range. Notably, in the Saddle Peak Hills, clasts identical in composition and facies to that of the Noonday Dolomite—a unit previously interpreted as the post‐Marinoan cap carbonate—are incorporated into diamictites at the top of the Kingston Peak Formation. Combined with the carbonate‐rich composition of rocks at the top of the formation, these observations suggest that the uppermost diamictites of the Saddle Peak Hills and Alexander Hills are genetically related to the Noonday Dolomite and are unrelated to glacial processes. We propose that they formed through local slope foundering and basinward collapse of the adjacent carbonate platform, substantiating recent interpretations of Noonday carbonate platform dynamics, and demonstrating that they are genetically unrelated to Cryogenian glaciation. Thus, clast textural analyses play a valuable role in establishing whether contested ‘snowball Earth’ outcrops are truly glaciogenic or simply the product of local slope collapse.


| Geological setting and stratigraphy
Cryogenian diamictites occur in the Kingston Peak Formation of the Panamint Range, to the west of Death Valley (Petterson, Prave, & Wernicke, 2011), and in a suite of disconnected outcrops to the southeast of Death Valley (Le Heron, Busfield, Ali, Vandyk, & Tofaif, 2019). Key glaciogenic indicators comprise faceted, striated clasts and delicate dropstone textures, with local evidence for glaciotectonic deformation (Busfield & Le Heron, 2016 and references therein). In an influential paper, Prave (1999) proposed a tectonostratigraphy for the Death Valley region that recognized two periods of glaciation with coeval rifting, linked to the Sturtian and Marinoan glaciations, and overlain by an Ediacaran cap carbonate, the Noonday Formation (Petterson et al., 2011). This framework was later adapted by Macdonald et al. (2013), who assigned the Marinoan glaciation to a locally developed mapping unit in the uppermost Kingston Peak Formation, named KP4, and assigned all underlying glaciogenic strata to the Sturtian glaciation. In the Death Valley area, many exposures contain interbedded diamictites and ironstones. On geochemical grounds, ironstones interbedded with the Kingston Peak Formation diamictites are now considered unlikely to relate to rift-related hydrothermal activity as proposed for other localities (e.g. Cox et al., 2016) but rather to the production of oxygenated brines at the palaeo-ice margin (Lechte, Wallace, van Smeerdijk Hood, & Planavsky, 2018). Macdonald et al. (2013) described and mapped KP4 in the Saddle Peak Hills, whereas in the Kingston Range they noted its presence but provided no description or indication of its location. In the uppermost Kingston Peak Formation of the southern Kingston Range, Le Heron, Busfield, Ali, Tofaif, and Vandyk (2018) described a diamictite, with striated clasts and stratified intervals containing lonestones, intercalated with turbidites. They interpreted these deposits, including the diamictites, collectively as representing the upper part of a F I G U R E 1 Location map of the Death Valley study region, showing the three study areas in this paper, namely (a) the southern Kingston Range, (b) the Alexander Hills, and (c) the Saddle Peak Hills. Each of these disparate study areas yields excellent but differing facies of the Kingston Peak Formation. Section (d) corresponds to the Valjean Hills, from which we present no data herein, but from where carbonate-rich diamictites at the top of the Kingston Peak Formation have previously been reported (Mrofka & Kennedy, 2011). Map from Lechte et al. (2018) a d subaqueous fan system in a proglacial setting that received ice-rafted sediment. This diamictite does not extend over the entirety of the southern Kingston Range and is locally absent (see fig. 5 of Lechte et al., 2018). In the Saddle Peak Hills, Creveling et al. (2016) described dropstone-bearing turbidites sharply overlain by polymict conglomerates (herein categorized as diamictites), marking the base of KP4. The clasts within these diamictites transition upwards from older Neoproterozoic lithologies of the underlying Pahrump Group, through clasts of the immediately underlying Kingston Peak Formation, to megaclasts (sensu Terry & Goff, 2014) of the Ediacaran Noonday Formation. Creveling et al. (2016) proposed two depositional scenarios, both invoking lowstand filling of submarine incised valleys either 'Scenario 1': KP4 was deposited prior to the Noonday Formation, supplied by glacial and/or alluvial processes. Importantly, KP4 in this interpretation only comprises the diamictite above the lonestone-bearing turbidites but below the first appearance of Noonday megaclasts. By contrast, the Noonday megaclastbearing diamictite rests disconformably upon KP4 and represents an intra-Noonday Formation lowstand; or 'Scenario 2': KP4 comprises the full thickness of diamictite, from immediately above the dropstone-bearing turbidites to the overlying dololaminites of the Noonday Formation, including the Noonday megaclasts. This whole diamictite then represents the same intra-Noonday Formation lowstand deposit.

| METHODOLOGY
The uppermost diamictites of the Kingston Peak Formation were studied in three Death Valley outcrop belts by the authors during their 2017 field season, focussing on three localities (Figure 1): (a) the uppermost diamictite in the Saddle Peak Hills, including KP4 as described by Macdonald et al. (2013) and the overlying Noondaymegaclast-bearing diamictite (i.e. 6-48 m of section SP01 of Creveling et al., 2016, Figure 4); (b) mapping unit PꞒK4 of Wright (1974) in the Alexander Hills (Figure 2), reassigned to KP4 by Mrofka and Kennedy (2011); (c) diamictites previously interpreted as 'glacial' in the uppermost Kingston Peak Formation of the Kingston Range . In each of these areas detailed laterally equivalent sedimentary logs were completed, in particular focussing on clast lithology and morphology. Additionally, multiple 1 m 2 quadrats were placed at different stratigraphic levels over the outcrop, from which both lateral and vertical variation in clast composition and roundness, using Powers (1953) classification, were described. Quadrats were subsequently photographed and in the Alexander Hills only the PꞒK4 unit was remapped. For the quadrat analysis, the focus was placed on clasts in the range ø-4 to ø-8 on the Udden-Wentworth scale (i.e. coarse gravel to boulder-scale). Megaclasts (i.e. clasts greater in F I G U R E 2 Geological map of the Alexander Hills, adapted from Wright (1974 size than ø-8; see Terry & Goff, 2014) were noted and described, but excluded from the quadrat analysis. The clast textural analysis was undertaken in the field (i.e. roundness and lithology were described at outcrop). Although recent work has highlighted the use of micromorphology/thin section analysis to describe and interpret Cryogenian diamictites , this methodology was considered unsuitable for our study sections. Owing to the high carbonate content of the outcrops in the Saddle Peak Hills and Alexander Hills, grain-boundary dissolution and stylolitization adversely affects textures in thin section.

| Saddle Peak Hills: Descriptions
In the western part of the Saddle Peak Hills the uppermost diamictite is up to 30 m thick and rests disconformably upon repetitively-bedded, ferruginous turbidites that bear occasional lonestones (Figures 3a, 4 and 5a). These latter rocks are very similar to the well-studied succession exposed a few kilometres to the east at Sperry Wash (Busfield & Le Heron, 2016). The diamictite expresses a well-developed coarsening upward, in terms of maximum clast size, with the majority of Noonday Formation megaclasts concentrated in the upper 10 m (Figures 4 through 6). It is conformably overlain by dololaminites of the Noonday Formation (Figure 5a-c). The uppermost part of the megaclast-bearing diamictite is dominated by lath-shaped, pencil-shaped and generally oblately-shaped dololaminite clasts of identical composition to the overlying, undisturbed, dololaminites (Figure 5b,c). Panoramic views of the megaclast-bearing diamictite ( Figure 5a) reveal that the megaclasts exhibit a spectrum of rounded, equant shaped to angular, slab-like morphologies. Some of the latter include sub-angular clasts representing part of a Bouma sequence resembling the underlying Kingston Peak Formation (Figures 3a and 6a). However, the majority of the megaclasts comprise carbonate (Figures 3a and 6b), some with highly distinctive features typical of cap carbonates (Creveling et al., 2016;Shields, 2005), including folded and deformed dololaminite clasts of identical appearance to the overlying Noonday Formation (Figure 6c: compare with Figure 5b) and pipe-like features ( Figure 6d). Within the quadrats, carbonate clasts ( Figure 5d) are dominant and comprise 98% of the 208 clasts recorded (Figure 7), although not all are identifiably of Noonday Formation origin, whereas sandstone ( Figure 5e) and mudstone clasts are rare. They also reveal that highly irregular clast shapes are common, with 92% classified as angular or very angular using Powers (1953) roundness scale (Figure 7c). Neither the quadrats nor measured sections (excluding megaclasts) suggest that there is an upward change in the dominance of carbonate clasts or their angular character.

| Alexander Hills: Descriptions
This region was mapped by Wright (1974), who divided the Kingston Peak Formation into mapping units pꞒk1-pꞒk4, later reassigned as KP1-KP4 by Mrofka and Kennedy (2011). The reader is referred to Le Heron et al. (2019) for a review of this issue. The map presented herein ( Figure 2) only slightly modifies the earlier work of Wright (1974). In detail, the boundary of the uppermost Kingston Peak Formation diamictite, pꞒk4, now forms a wedge-shaped geometry that pinches out to the north and terminates against the Sheephead fault to the south. This unit comprises megaclast-bearing diamictite in which multimetre-sized clasts are set within a 'matrix' of pebbly to boulder-rich diamictite (Figures 8d,e and 9). It occurs immediately above graded sandstones of pꞒk3 (Figure 8c), very similar to those beneath KP4 of the Saddle peak Hills, although the contact between these units is typically concealed. On Figure 8d, for example, the geologist is standing on ferruginous graded beds (pꞒk3), exposure immediately above is patchy and large orange and brown megaclasts (pꞒk4) up to 6 m in diameter crop out further up the cliff section. Beneath these sandstones are boulder-bearing diamictites of unit pꞒk2, within which clasts are more rounded than pꞒk4 and no megaclasts occur (compare Figures 8b and 9). Panoramic photographs illustrate that some of the megaclasts are arranged as trains of multi-metre scale blocks, for example four brown megaclasts left of 'E' in Figure 8a. These are encased within a predominantly silty, red-coloured, and highly heterogeneous matrix (Figure 8e).
In the Alexander Hills, the quadrat analyses reveal that highly irregular boulder and cobble clasts are commonplace, representing the 'matrix' to the megaclast-bearing diamictite. There are several examples of adjacent clasts arranged like the pieces of a jigsaw puzzle (Figures 3b, log ii and 9a). In terms of clast count (Figure 9b), whilst broadly comparable to the Saddle Peak Hills in terms of a dominant carbonate clast composition (93%), there is a slightly greater proportion of sandstone clasts (6%). Using Powers (1953) index, some 76% of studied clasts are either angular or very angular, with only 1% described as rounded ( Figure 8c). As in the Saddle Peak Hills, there is no clear stratigraphic trend in these characteristics.

| Southern Kingston Range: Descriptions
The southern Kingston Range has been the subject of considerable recent detailed investigation, including detailed sedimentary logging through the full 2.5 km thickness of the Kingston Peak Formation . In a similar manner to the Saddle Peak and Alexander Hills areas, a thick (40 m) diamictite-dominated package appears in the   stratigraphy immediately below the Ediacaran Noonday Dolomite and above a series of graded beds ( Figure 3c). This diamictite is punctuated by lenticular bedsets of graded conglomerate, graded sandstone and mudstone ( Figure 10a). Sharp contacts separate the typically brown-red graded beds from the typically green diamictites (Figure 10b). Unlike the Saddle Peak and Alexander Hills, these diamictites vary between massive and stratified varieties (Figure 10c). Excellent examples of lonestones occur both in mudstones found between graded sandstone beds ( Figure 10d) and within the stratified diamictites ( Figure 10e). Quadrat analysis of the upper diamictite in the Kingston Range ( Figure 11) illustrates that the clast populations are more heterogeneous than their counterparts in the Saddle Peak or Alexander Hills. Particularly notable are large clasts of metabasite in the study area ( Figure 11a): very similar to metabasites forming megaclasts further down the succession (Le Heron et al., 2018), derived from a 1.08 Ga diabase intrusion into the Crystal Spring Formation (Calzia et al., 2000;Heaman & Grotzinger, 1992;Vandyk et al., 2018). In terms of clast populations, whilst dolostone is the dominant lithology (62%) and sandstone accounts for almost a quarter (23%) of described clasts, metamorphic clasts (schist and gneiss) make up 8% of the 313 counted clasts (Figure 11b). By further contrast with the Saddle Peak and Alexander Hills sections, 73% of clasts classify as rounded and well-rounded on Powers (1953) scale: only 2% are angular and there were no very angular clasts observed (Figure 10c).

| INTERPRETATIONS
In the Saddle Peak Hills, diamictites and their clasts form the 'matrix' in which megaclasts of the Kingston Peak and Noonday formations occur. Despite the restriction of Noonday megaclasts to the upper third of these diamictites, the quadrats reveal that there is no clear up section change in the clasts within this 'matrix'-carbonate clasts remain dominant at all stratigraphic levels, as does their strikingly angular character. The concentration of Noonday megaclasts in the upper 10 m suggests dispersive pressures within a debris flow, possibly in concert with kinetic sieving (Creveling et al., 2016). The lath-shaped, pencil-shaped and oblately-shaped dololaminite clasts testify to very low transport distances and most likely F I G U R E 4 Panoramic overview of the Saddle Peak Hills section, showing the stratigraphic relationships between repeatedly stacked, redcoloured turbidites below, a diamictite with megaclasts in the middle, and the Noonday Formation capping the poorly sorted rocks at the top F I G U R E 3 A suite of sedimentary logs from each of the three study areas in this paper. Since this paper focusses on distinguishing glacial versus non-glacial diamictites in the upper part of the Kingston Peak Formation, only sections relevant to the upper diamictite are shown. (a) Two measured sections from the western side of the Saddle Peak Hills, where the contact between the Kingston Peak Formation and overlying cap carbonates of the Noonday Dolomite is excellently exposed. (b) A series of five sedimentary logs from the Alexander Hills which illustrate progressive thickening of the megaclast-bearing diamictite from left to right, i.e. from north to south. The positions of these are shown both in map view on Figure  steep slopes at a basin margin. The occurrence of angular blocks of material, derived from the graded beds immediately beneath the 'KP4' unit, suggests that slope failure involved the Kingston Peak Formation as well as the evolving carbonate platform. The folding of these blocks is interpreted as having occurred in soft sediment, suggesting only partial lithification. This might imply that only a limited period of time separated deposition of the glaciogenic turbidites of the Kingston Peak Formation and the overlying carbonate deposits of the Noonday Formation. It is worth noting that Creveling et al. (2016) envisaged upsection changes in the composition of the 'KP4' unit based on megaclasts. No clear upsection changes in 'KP4' are recognized herein, and yet this does not necessarily contradict their findings. This is because the quadrat analysis employed here focussed on material with a maximum size of ø-8 (boulders), corresponding to the 'matrix' of a megaclast-bearing diamictite.
In the Alexander Hills, it can be speculated that because the thickest occurrence of the pꞒk4 succession abuts against the major NW-SE striking Sheephead Fault and pinches  (Walker, Klepacki, & Burchfiel, 1986): a major olistostrome complex was shed from these extensional fault arrays (Le Heron, Busfield, & Prave, 2014). Likewise, pronounced lateral thickness variations in the Kingston Peak Formation of the Panamint Range attest to syn-depositional tectonics (Miller, 1985). Thus, there is a precedent of 'sudden' appearance of megaclast-bearing material, and associated dramatic lateral thickness changes, adjacent to major fault systems in Death Valley, which demonstrates a structural control on sedimentation. It therefore appears that the E-W striking fault crosscutting the outcrop was active during Kingston Peak Formation time. Texturally, all evidence also points to a local origin via a collapsing carbonate platform, in a similar manner to the Saddle Peak Hills sections, rather than to a possible glacial deposit (Prave, 1999). Whilst sandstone clasts are common, dolostone clasts remain dominant in the quadrat analysis performed here, and the predominance of angular to very angular clasts suggests, by direct comparison to the Saddle Peak Hills (Creveling et al., 2016), a very local source, probably as a result of toppling of material over a fault scarp into the basin. It is argued here, therefore, that the pꞒk4 unit is 'an olistostrome in miniature', implying local derivation and fault control, by reference to studies of the Kingston Peak Formation in neighbouring ranges (Le Heron et al., 2014. By contrast with the Saddle Peak and Alexander Hills rocks, which are interpreted as unrelated to glaciation, the diamictites at the top of the Kingston Peak Formation in the southern Kingston Range are interpreted as glacially derived. The arguments revolve around (a) the presence of (2018) point out the occurrence of lonestones that downwarp laminations in stratified diamictites, and thus posit an origin via ice-rafting. The massive diamictites, by contrast, were viewed by those authors as a refluxed variety of this facies (glaciogenic debris flows). Additional evidence for a glacial origin lies in the occurrence of striated clasts, together with the much broader range of clast lithologies. The agent for transporting clasts over long distances was, in the case of the upper diamictite in the southern Kingston Range, the glacial conveyor belt. There is one aspect to the interpretations which stands in stark contrast with general assumptions about textures predicted in ancient glacial sediments. The observation that some 73% of the clasts studied in the Kingston Range quadrat analysis are rounded to wellrounded is perhaps counter-intuitive, but can be understood in terms of recycling of clasts both through englacial, subglacial and proglacial processes. In other glacial records, rounded but beautifully striated clasts are very commonplace because they record several generations of recycling (Tofaif, Le Heron, & Melvin, 2019).

| DISCUSSION
In each of the Death Valley outcrop belts studied, the uppermost Kingston Peak Formation consistently comprises graded beds overlain by diamictite, capped by dololaminites of the Noonday Formation. However, it is possible to argue that these similarities in general facies, stratigraphic context and stratigraphic position are in themselves insufficient to assign tectonostratigraphic or chronostratigraphic significance to these units. Neither in the Saddle Peak nor in the Alexander Hills does the interpretation presented here suggest that the uppermost diamictites are glaciogenic. Conversely, in the southern Kingston Range there is support for the previous interpretation of a glaciogenic affinity .
Despite the operation of comparable depositional processes in the Saddle Peak and Alexander Hills, coupled with apparently identical stratigraphic context and position, caution should be employed against any assumption that the two were deposited coevally. The idea that megaclastbearing diamictites are locally derived and result from  Figure 2. Approximately 100 m to the right of the photograph (i.e. to the south), a major NW-SE striking fault-the Sheephead Fault-cuts off the Kingston Peak Formation, juxtaposing it against the Quaternary China Ranch Beds (Wright, 1974). (b) Detail of the lower diamictite in the Alexander Hills, corresponding to pꞒk2 of Wright (1974). (c) Graded sandstones dipping steeply to the right of the photograph, overlain by a thin diamictite on which the lens cap sits, pꞒk3 of Wright (1974).  Macdonald et al., 2013;Prave, 1999;Walker et al., 1986). During rifting, time-transgressive evolution of fault systems (Alves et al., 2009;Eyles & Januszczak, 2004) would be expected at the margins of the Death Valley basin, and thus correlating deposits laid down in the lee of degrading fault scarps is fraught with difficulty. Nevertheless, there is one scenario wherein the locally derived megaclastbearing diamictites of the Saddle Peak and Alexander Hills might be broadly time-equivalent and genetically related.
In the central Kingston Range, Le Heron et al. (2014) posited that a major olistostrome complex in the middle of the Kingston Peak Formation may have been emplaced during interglacial isostatic rebound. Glacioisostatic rebound was also entertained as a viable scenario by Creveling et al. (2016), for the uppermost diamictites in the Saddle Peak Hills, and it is proposed that this mechanism can extend to the pꞒk4 unit in the Alexander Hills. However, in this model, pre-existing faults may have been reactivated during or following retreat of the Kingston Peak Formation ice sheets, in a manner similar to the 'piano-key tectonics' envisaged for Pleistocene glaciated basins (Eyles & McCabe, 1989). It is therefore more likely that rebound-related deposits were diachronous, underscoring the argument that they have  (Mahon et al., 2014;Vandyk et al., 2018). In the Valjean Hills, Mrofka and Kennedy (2011, p. 453) also noted that 'Noonday Dolomite clasts are included in diamictite of the KP4 member or are in diamictite interbedded with KP4 member sedimentary breccia', suggesting that the model proposed herein can also be extended to those sections, although further work to establish this is required. The uppermost diamictites of the Saddle Peak and Alexander Hills, with their megaclasts derived from slope failure, can be considered modest examples of olistostromes. In this context, their diachronous emplacement by slopefailure is well precedented within stratigraphically lower parts of the Kingston Peak Formation. For example, one olistostrome occurs in both the Goler Wash of the Panamint Range (Prave, 1999) as well as in the central and southern Kingston Range (Figure 1; Calzia et al., 2000;Le Heron et al., 2014;Macdonald et al., 2013). The southernmost Kingston Range exposes by far the thickest Kingston Peak Formation outcrop in the entire Death Valley area, and would presumably therefore be expected to contain the most complete sedimentary archive . Despite this, the Silurian Hills succession is thinner yet has four olistostrome intervals, generally <300 m thick, with megaclasts of both carbonate and metabasite (Le Heron et al., 2017). Regionally, the different number of olistostrome intervals strongly suggests that basin-bounding faults operated diachronously from outcrop belt to outcrop belt, undermining their use as chronostratigraphic markers.
Le Heron et al. (2017) noted the angular character and dominance of locally derived clast-types in the Silurian Hills and was able to use these criteria to distinguish slope-failure from glacial deposit. This same difference in clast composition and angularity also distinguishes the uppermost diamictites of the Saddle Peak and Alexander Hills from those of glacially derived diamictites of the Kingston Range. Importantly, this suggests that these criteria may be used to distinguish slope-failure from glaciogenic diamictites elsewhere in the region. The remainder of this section uses this concept to consider whether the Marinoan glaciation is represented in the Panamint Range and therefore in the Death Valley region at all. The two-glaciation stratigraphic framework established in the Death Valley region (Miller, 1983(Miller, , 1985Prave, 1999) has played an important part in the subsequently developed tectono-stratigraphy of the SW Laurentian margin (Macdonald et al., 2013;Yonkee et al., 2014). It originated in the Panamint Range, where the uppermost diamictite-the Wildrose submember (Miller, 1987)-has been interpreted as representing the Marinoan glaciation on account of its erosional, unconformable base and position above a Sturtian cap carbonate (Sourdough Limestone; Macdonald et al., 2013;Prave, 1999). However, the problem of distinguishing between Ediacaran slope-failure and glaciogenic debris flows is equally pertinent for these strata as it is for the Saddle Peak Hills (Creveling et al., 2016;Miller, 1987). The Wildrose diamictite shares key features with the slope-failure diamictites of the Saddle Peak and Alexander Hills (Miller, 1987): (a) a down-cutting erosional base, reworking underlying material; (b) carbonate megaclasts up to 3 m in length, including angular clasts resembling the Noonday Formation; (c) structureless diamictite, with rare clast trains; (d) uneven distribution, only locally developed; and (e) complete lack of direct glacial indicators such as striated clasts, dropstones or subglacial deformation. Added to this there is the suggestion that this diamictite interfingers with the Noonday Formation, precluding a glaciogenic interpretation (Miller, 1987) (Miller, 1987, Figures 2 and 6). This similarity to the Saddle peak and Alexander Hills, combined with its stratigraphic context, might suggest that the Wildrose diamictite represents carbonate platform collapse and that the Marinoan glaciation is absent from the Death Valley region. Miller (1983) provided clast-count data from 23 different examples of the Wildrose diamictite, which, by comparison to the data presented herein, can be used to test this hypothesis.
The compositional heterogeneity of clasts from the Wildrose diamictite ( Figure 12) draws closer comparison to the Kingston Range uppermost diamictites, than the slope-derived deposits of the Saddle Peak Hills or the Alexander Hills. Carbonate clasts are common (34%), but quartzite is dominant (40%), and a range of other clast types (granites: 13%; gneiss: 3%; diabase: 1%) are also present. There is considerable variation between samples of the same section, for example at Goler Wash quartzite clasts vary from 28% to 81% (GW592, GW594), which is a feature of the Kingston Range but not the Saddle Peak or Alexander Hills diamicites (cf. Figures 7, 9 and 11). These clast-count data support the suggestion of Prave (1999) that the carbonate with which the Wildrose diamictite interfingers is in fact the Cryogenian 'un-named' limestone and not the Noonday Formation. Prave (1999) noted differences in lithology between the un-named limestone and the Noonday Formation in the Surprise Canyon, despite Miller (1983Miller ( , 1985Miller ( , 1987 stating they are alike. Furthermore, the carbon isotope values of the un-named limestone, around 6‰ (n = 3), is unusually heavy for a cap carbonate. In summary, a non-glaciogenic slopefailure source cannot be ruled out for the Wildrose submember. However, if the clast compositional data are interpreted similarly to the sections from the Saddle Peak Hills and the Alexander Hills, then a glacial origin also remains possible. Further investigation is required to resolve this question.

| CONCLUSIONS
Reappraisal of diamictites in the uppermost part of the Kingston Peak Formation, Death Valley, reveals that F I G U R E 1 2 Pie chart of clast count data from 23 samples of the Wildrose submember in Miller (1983 (Miller 1983) some deposits previously allied to the Marinoan glaciation are non-glacial in character. These deposits, which are exposed in the Saddle Peak and Alexander Hills, represent material that is derived from the immediate area and was produced through carbonate platform collapse events. Clast counting, not including megaclasts, reveals a dominant (>90%) dolostone component, with the vast majority of clasts classified as angular to very angular on Powers (1953) roundness chart, which does not change up-section. In the Saddle Peak Hills, this lack of change reinforces the idea that the KP4 diamictite of Macdonald et al. (2013) is genetically related to overlying diamictite containing megaclasts of foundered Noonday Formation cap carbonate. In the Alexander Hills, there is equally little reason to interpret the uppermost diamictite as glacially derived. Nevertheless, the likely diachronous nature of both rifting and/or deglacial uplift conspire to prevent chronostratigraphic links being made between the uppermost Saddle Peak and Alexander Hills diamictites. By contrast, analysis of a diamictite section at a comparable stratigraphic level in the southern Kingston Range reveals that glacially derived deposits differ greatly in terms of their outcrop style, associated lithofacies, clast composition and fabric. In this comparative section, the presence of dropstones and striated clasts in graded interbeds and stratified diamictites, underscores their glacial affinity. In the latter area, (a) the greater variety of clast-types is supportive of a wide-ranging provenance area typical of glacially transported debris, and (b) the greater degree of roundness of the clast population is suggestive of several generations of reworking rather than local slope collapse followed by deposition. These clast characteristics are shared by the Wildrose submember, which purportedly represents the Marinoan glaciation in the Panamint Range, west of Death Valley. On this basis, despite its similarities in general facies, stratigraphic context and stratigraphic position to the slope-deposits of the Saddle Peak and Alexander Hills, a glacial interpretation is supported for the Wildrose diamictite. It has been demonstrated herein that coincidence of stratigraphic context and general facies type between outcrops is, in the case of diamictites, insufficient to demonstrate either coeval deposition or a glacial affinity. Instead alternative lines of evidence, such as clast-count data, should be used to verify such assumptions, which in this case have cast considerable doubt as to whether the Marinoan glaciation may be recognized in the SE Death Valley region.