Glass shards from the A.D. 1783 Laki fissure eruption in Iceland have been identified in the GISP2 ice core from Summit, Greenland, at a level just preceding the major acidity/sulfate peak. Detailed reconstruction of ice stratigraphy, coupled with analyses of solid particles from filtered samples, indicate that a small amount of Laki ash was carried via atmospheric transport to Greenland in the summer of 1783, whereas the main aerosol precipitation occurred in the summer and early fall of 1784. Sulfate concentrations in the ice increase slightly during late summer and fall of 1783 and remain steady throughout the winter due to slow oxidation rates during this season in the Arctic. The sulfate concentration rises dramatically in the spring and summer of 1784, producing a massive sulfate peak, previously believed to have accumulated during the summer of 1783 and commonly used as the marker horizon in Greenland ice core studies. The chronology of ash and acid fallout at the GISP2 site suggests that a significant portion of the Laid eruption plume penetrated the tropopause and that aerosol generated from it remained aloft for at least 1 yr after the eruption. Based on comparisons with other glaciochemical seasonal indicators, abnormally cool conditions prevailed at Summit during the summer of 1784. This further supports the claim that a significant volume of sulfate aerosol remained in the Arctic middle atmosphere well after the eruption had ceased.

Devils Hole, in southern Nevada, is a surface collapse into a deep, planar, steeply dipping fault-controlled fissure in Cambrian limestone and dolostone. The collapse intersects the water table about 15 m below land surface and the fissure extends at least 130 m deeper. Below water, most of the fissure is lined with a >30-cm-thick layer of dense maxillary calcite that precipitated continuously from groundwater for >500,000 yr. The thick mammillary calcite coat implies a long history of calcite-supersaturated groundwaters, which, combined with the absence of dissolutional morphologies, suggests that Devils Hole was not formed by karst processes. Devils Hole is located in a region of active extension; its tectonic origin is shown by evidence of spreading of its planar opening along a fault and by the orientation of its opening and others nearby, perpendicular to the northwest-southeast minimum principal stress direction of the region. Most Quaternary tectonic activity in the area, including seismicity and Quaternary faults and fractures, occurs on or parallel to northeast-striking structures. The hydrogeologic implications of this primarily structural origin are that fracture networks and caves opened by extensional tectonism can act as groundwater flowpaths functionally similar to those developed by karst processes and that, during active extension, transmissivity can be maintained despite infilling by mineral precipitation. Such extensional environments can provide conditions favorable for accumulation of deposits preserving long, continuous paleoenvironmental records. The precipitates in Devils Hole store chronologies of flow system water-level fluctuations, hydrochemistry, a half-million-yr proxy paleoclimate record, evidence of Devils Hole's tectonic origin, and probably atmospheric circulation.

In July 1988, the FlatFish Range Fire burned over the type-Pinedale moraines at Fremont Lake, Wyoming, and caused extensive exfoliation of exposed boulder surfaces. The mass of exfoliated material from 130 of 1030 boulders investigated was measured and recorded with information concerning factors that could influence the extent of fire-induced exfoliation. The range in thickness of material removed from 98 randomly selected boulders within the burn area (averaged over the entire exposed boulder surface area) is 6.1 to < 0.1 mm. The mean thickness loss for all 98 boulders is 0.9 mm/fire and the expected loss from individual boulders (median) is 0.4 mm/fire. At the 95% confidence level there is no significant relationship between the degree of exfoliation and boulder size, lithology, grain size, proximity to vegetation, or vegetation density. The expected fire-induced boulder surface erosion rates range from 5.9 to 0.3 x 10-3 mm/yr on boulders in sagebrush rangeland where fire recurrence intervals are typically every 20 to 400 yr. Fire-induced exfoliation may account for differences in boulder size and abundance on Pinedale and Bull Lake moraines. Surface dating methods using varnish or cosmogenic nuclides may yield exposure ages that are too young if the consequences of range fires are not considered when sampling boulder surfaces that are within about 2 m above ground level.

In the northern Owens Valley, the Owens River has cut and preserved three terraces uplifted above its modern floodplain. We have analyzed soils and used other relative-dating techniques in order to differentiate the terrace surfaces, to correlate from the terraces to moraines of the eastern Sierra Nevada, and to test the effects of relief and climatic differences on relative-dating results. Most of the relative-dating techniques attempted were successful in differentiating the terraces of the Owens River sequence. In addition, weathering rinds, soil descriptions, and an early Holocene 14 C date for the youngest terrace support correlation of the three terrace levels with the three post-Sherwin-age glacial stages (Tioga, Tahoe, and pre-Tahoe) found in nearby canyons of the Sierra Nevada. Other relative-dating techniques that are commonly employed, however, suggest distinct differences between the terrace and moraine sequences. These differences probably reflect the difference in relief, such that slow erosion of the moraine crests may have retarded soil development relative to the flat terrace surfaces and altered surface-weathering processes. Use of relative-dating analyses to correlate between dissimilar geomorphic systems requires caution but use of a broad range of techniques and absolute-age calibration may make correlation possible.

The geomagnetic secular variation record retained by glaciolacustrine and marine sediments at nine sites in northern New York and southern Ontario provides a means for stratigraphic correlation of glacial deposits for the time period between about 12,600 to 9900 14C yr B.P. Measurement of the depositional remanent magnetism of sediments deposited in Glacial Lake Iroquois and the Champlain Sea has produced a geomagnetic secular variation curve that represents the time period immediately following deglaciation about 12,600 14 C yr B.P. The curve varies from about 358° to 344° declination and 51° to 61° inclination over approximately 180 valve years. Marine sediments of the Champlain Sea have preserved a record approximately 1500 yr long that varies from about 2° to 29° declination and 47° to 60° inclination. These combined glacial-paleomagnetic records may also correlate with those from glacial sequences beyond our study area. The shape and amplitude of the secular variation record in glaciolacustrine and marine sediments from the western Adirondack borderland show agreement with other glacial varve secular variation records and suggest possible correlations with secular variation curves from lake cores.

Dendrochronological analysis of fossil wood from Two Creeks, Wisconsin, reveals that the Two Creekan Intetstade lasted at least 252 yr. The sites crossdated by tree rings cover an area of about 970 km2. AMS determinations from the beginning and end of the chronology open a 14 C time window for the episode from 12,050 to 11,750 yr B.P. The interval is contemporaneous with the Older Dryas in northern Europe. The development of a forest covering at least 970 km2 on the western shore of Lake Michigan indicates a water level about as low as in modern times. Glacier retreat must have opened drainage channels either through the Straits of Mackinac or via the Indian River Plateau into the eastern lakes. The beginning of the tree-ring chronology coincides with the peak of meltwater pulse 1A at 12,000 yr B.P. Increased amounts of meltwater seem to have disturbed the heat exchange between the waters and the atmosphere in the North Atlantic off the Gulf of St. Lawrence or affected the δ18O-ratio of the evaporation, causing the climatic or isotopic reversal of the Older Dryas in Greenland and northern Europe.

Radiocarbon ages from the bottom peat deposits and detailed morphometric measurements at five peatlands in the boreal region of southern Finland were used to reconstruct the Holocene patterns of raised mire development. The results suggest that the areal expansion of the mires has been extremely rapid at times—up to several hundreds hectares per millennium—but the periods of more rapid expansion were not coincident among the mires. The rate of lateral extension was shown to depend mainly on the topography of the mineral substrate. Horizontal expansion did not always correlate with vertical growth. Mires expanded slowly between ca. 6000 and 4000 yr B.P., probably as a consequence of drier climatic conditions at that time. The mires also expanded slowly during the two to three recent millennia, the reason for this most probably lying in steepening of the bottom gradient toward the marginal areas of the present mires, The results point to a substantial measure of local environmental control over the lateral extension in raised mire ecosystems.

The Abric Romani continental pollen record provides evidence of abrupt climate variations during the last glaciation. The pollen record reveals a pattern of high-frequency variations similar to that in ice cores as shown by oxygen isotopes or dust content. Analyses of 14 travertine samples yield coherent U-series ages in correct stratigraphic order, ranging from 40,000 to 70,000 yr. According to changes in the composition and structure of the vegetation, five paleoclimatic phases have been differentiated; the oldest, ranging from 70,200 to 65,500 yr ago, records thermophilous taxa which correspond to the last warm events of isotopic stage 5. The next phase, from 65,500 to 56,800 yr, records a cold, humid climate that correlates with isotope stage 4. During the third phase, from 56,800 to 49,500 yr, relatively warm and cold events alternate. The fourth phase, from 49,500 to 46,200 yr, records a cold, dry climate. The fifth phase, ranging from 46,200 to 40,800 yr, records a milder phase that is similar in appearance to the beginning of the Holocene.

Pollen, plant macrofossil, chironomid, and lithological analyses of late-glacial sediment from Pine Ridge Pond all indicate that the warming following deglaciation was punctuated on two separate occasions by a return to colder climes; initially by a brief and moderate cold episode (the Killarney Oscillation), followed shortly by the Younger Dryas. The Killarney Oscillation, AMS 14C dated at between 11,290 and 10,960 yr B.P., is characterized by abrupt and simultaneous increases in Betula (birch) pollen and Betula glandulosa (dwarf birch) macrofossils and a decline in tree pollen. The Younger Dryas is distinguished by the reversion of a short-lived closed Picea (spruce) forest to a prolonged forest-tundra stage (ca. 10, 820 to 9500(?) yr B.P.), during which macrofossils of arctic/alpine plants such as Salix herbaces, Cassiope hypnoides, Armeria maritima, and Thalictrum alpinum make their sole late-glacial appearance. Chironomid-inferred summer surface-water temperatures indicate a cooling of approximately 8 and 12°C during the Killarney Oscillation and Younger Dryas, respectively, and coarse-sand analysis indicates increased erosion on the landscape during these same intervals.

The mammalian assemblage and archaeological finds from the Lower Palaeolithic hominid site of Evron Quarry, situated on the northern coastal plain of Israel, are described and discussed. In their lithic and faunal composition, the sites of Latamne (QfIII) (Latamne Formation, Orontes, Syria) and Sitt Markho (Nahr elKebir, Syria) resemble Evron and are probably contemporaneous. It is suggested here, based on their lithic and faunal composition, that these sites may be chronologically closer to the site of Ubeidiya ('Ubeidiya Formation, Jordan Valley, Israel; 1.4 myr) than to the sites of Gesher Benot Ya'akov (dated as <800,000 yr B.P.), which differs in both aspects from Evron. The mammalian faunule from Evron comprises a biogeographical mixture, a result of biotic exchanges with Africa, the Oriental region, and the Palaearctic. This exchange may have been associated with a post-'Ubeidiya hominid dispersal, either from Africa or south Asia via the Levantine "corridor."

Variations in the proportions of the various groups of Rodentia collected from eight stratigraphic levels at Tighenif had previously been interpreted as representing environmental (climatic) changes. However, the processes involved in the formation of assemblages are still not well understood. Examination of fragmentation, crystallinity, and chemical composition of fossil bones and teeth, compared to recent samples, shows that a possible explanation for the variations in frequencies is the differential preservation of skeletal remains. Thus, the evaluation of postmortem effects is essential in determining the accuracy of the fossil record. The abundance and composition of remains are governed by the complex interaction of many factors. As indicated by actualistic studies, direct deductions from fauna lists must be avoided as much as possible.

The seismic refraction reversed profiling technique has been used to investigate the topography of the last interglacial soil (paleosol S1) within the central Chinese Loess Plateau near Xifeng. The results suggest an essentially flat-lying soil at a depth which varies by only a few meters over an area of more than 10 km2. In addition, the results indicate a high-velocity layer at 50-60 m depth which is thought to coincide with a layer of carbonate concretions at the base of paleosol S5. The results agree well with the local loess-paleosol stratigraphy for this area and indicate that the seismic refraction method is a rapid technique for investigating paleotopography.