Wednesday, July 20, 1994 METEORITE SEARCHES 1:45 - 3:45 p.m. Chair(s): M. M. Grady Harvey R. P.* Current Research Activities of ANSMET: 1. Recent Studies in the Walcott Neve Region; 2. Planned Future Activities 1. RECENT STUDIES IN THE WALCOTT NEVE REGION The 1993-94 Antarctic field season to the Walcott Neve region was the 17th overall for ANSMET (the Antarctic Search for Meteorites project, funded by the Office of Polar Programs of the United States National Science Foundation). 1993-94 marked the third season led by the author, the 18th season for co-PI W. A. Cassidy, and the 12th season for co-I J. W. Schutt, who served as mountaineer and safety officer. The Walcott Neve region has been a prolific source of meteorites for ANSMET field parties, including well-known meteorite sources such as the Lewis Cliff ice tongue and the MacAlpine Hills icefields. This season's research efforts were two-fold. The goal during the first part of the season was to collect ice samples and radio echo-sounding data from the Lewis Cliff ice tongue, in an effort to understand the glaciology of that stranding surface. The second was to systematically search for meteorites in an area informally called "Foggy Bottom", a set of unnamed nunataks at the southeast end of the Walcott Neve that had been visited previously for reconnaisance purposes and had not been systematically searched. During the first 10 days of the season, radio echo-sounding traverses down the length of the Lewis Cliff ice tongue and across it were performed to develop a crude three-dimensional model of the basement below the stranding surface. In addition, many previously installed surveying stations were re-surveyed in support of continuing ice- movement studies of the ice tongue. Finally, near the northern end of the ice-tongue, exposed volcanic dust bands suggest a tilted stratigraphic sequence in the ice. A 50-m channel sample was cut at the ice surface, perpendicular to the suspected time-stratigraphic sequence. This sample has been subdivided and distributed to various ice chemistry specialists, with the hopes of identifying the specific time sequence exposed by comparison to established ice chemistries from core studies. At the completion of this glaciology research, the field camp was relocated approximately 50 km south to the Foggy Bottom area. Reconnaissance and a short period of systematic searching during two previous seasons had established the presence of meteorites in this region. Roughly 100 meteorites had been recovered in this area, and we estimated another 200-300 might be present, based on similarities to other fields. We were pleasantly surprised, therefore, to find an abundance of meteorites at various localities in the are. Many of these meteorites were found scattered among terrestrial rocks in moraines and on firn near the edges of icefields. The vast numbers of meteorites did not allow us to complete systematic searching of the area, and thus we are compelled to revisit the region in the near future. A total of 858 meteorites were recovered during the 1993-94 ANSMET season. The vast majority of these are ordinary chondrites, and particularly may represent a small number of shower falls rather than a large number of individual falls. However, the several achondrites, carbonaceous chondrites and metal-rich meteorites recovered should prove to be of significant interest to scientists. The most important meteorite find of the season was a single lunar specimen, similar to those found at MacAlpine Hills about 70 km away. The recovery of Antarctic meteorites continues to stimulate meteoritical research. As of April 1994, 8813 samples of Antarctic meteorites have been distributed to 260 researchers from 20 nations. 7078 meteorites have been recovered by ANSMET since its inception in 1976. Interest in Antarctic meteorites continues to be high and nothing is foreseen that would abate the world's desire for more meteorites. As of March 1994, over 55 men and women from 45 different institutions in 16 nations have participated as ANSMET field party members. 2. DIRECTIONS FOR FUTURE WORK. Although there are many theories as to why meteorite stranding surfaces exist and how they work to concentrate extraterrestrial specimens, our understanding remains fragmentary. There are many reasons for this; while each stranding surface appears to share a few broad traits, such as high ablation rates and sub-ice obstruction to ice flow, studies of individual icefields continue to frustrate us with their unique and complex characteristics. Recognizing the complex nature of the problem, we are making efforts to involve scientist from related disciplines in our studies of stranding surfaces. Our recognition of the complexity of meteorite stranding surfaces has also renewed our efforts towards setting up rigorous long-term experiments. During the coming field season we plan to establish a highly controlled, large-area ice movement and ablation network throughout the Foggy Bottom region. This will provide a superb framework for future studies by providing an accurate baseline of local ice flow and removal characteristics. We hope that this first step will encourage other scientists to work with us towards establishing an accurate recent history of the ice sheet for this important area near the head of the Beardmore glacier. As noted above, the 1994-95 ANSMET field party will return to the Foggy Bottom area, with a group of 6. This will be the fourth visit to the area, which has yielded around 1000 meteorites so far. We estimate that there may be another 200-300 meteorites in this region. If time allows, we will also search for meteorites on the nearby Goodwin nunataks icefield, and we hope to traverse to the MacAlpine Hills region where we can complete previous systematic searching efforts. Late in the season, twin otter aircraft will be used during reconnaissance of several icefields in the region of the Transantarctic Mountains lying between the Darwin and Byrd glaciers. Previously identified icefields in this region will be searched in detail for the first time since they were first visited in 1977. A two-person party will explore this region for approximately one week, in the hopes of estimating the requirements for future systematic searches. The 1995-96 season will likely visit other previously identified icefields further to the south (in the Grosvenor Mts. and Dominion Range region) via Twin Otter aircraft. Further in the future are plans for possible joint operations with the Australian Antarctic Division in 1996-97. ACKNOWLEDGEMENTS: This work supported by NSF grant OPP 91-175-58. Folco L.* Franchi I. A. Mellini M. Pillinger C. T. 1993/94 Antarctic Field Season. A Report on Activities Undertaken by the EUROMET/PNRA Meteorite Collection Expedition to Frontier Mountain, North Victoria Land Introduction This is a report on activities undertaken by the 1993/94 EUROMET/PNRA meteorite collection expedition to Frontier Mountain (FM), North Victoria Land, Antarctica, an area already recognised as a meteorite trap on the basis of previous finds by the 1984/85 GANOVEX IV and 1990/91 EUROMET/PNRA field campaigns [1,2]. The project, carried out within the framework of the IX Antarctic Campaign of the Italian Programma Nazionale delle Ricerche in Antartide (PNRA), foresaw two main objectives: i) to complete the collection of meteorites in the known productive sites and extend the systematic search into unexplored areas; ii) to initiate a thorough study of the meteorite concentration mechanism. A field team of five (L. Folco, I.A. Franchi, A.M. Fioretti, M. Meneghel and L. Boi) took part in this expedition operating from a camp downstream of FM, circa 3.5 km NE of the outcrop (72 degrees 57'20"S - 160 degrees 29'04"E), along the northern edge of the blue- ice field (Fig. 1), from 22/12/94 until 9/1/94. Systematic Search for Meteorites: Activities and Results A systematic search for meteorites was undertaken, both on foot and with skidoos, covering the entire blue ice field and all the local moraines, with the ultimate aim of studying the distribution of finds. Despite bad weather conditions and a widespread snow cover (the blue ice field was initially covered by a continuous bed of snow, which was reduced to about 50% after 10 days by strong winds), the search yielded a further fifty nine meteorite samples, weighing a total of about 4.5 kg. Meteorites were mainly found in the two previously discovered concentration sites [1,2]. The first, a trap for meteorites of suspected aeolian origin, is located circa 3.5 km due E of FM on the upwind slope of a morphological depression in the ice, locally trending EW (Fig. 1). The second is the ice-cored moraine of a valley (unofficially called "Meteorite Valley") in the southern sector of FM. Noteworthy is the recovery of three large samples from 658 g up to 1670 g. These finds, along with a sample weighing 942.3 g found in the 1984/85 field season [2], identify the strip of blue ice near the southern portion of FM as an accumulation site for particularly large masses. An additional five specimens were recovered in a wind-scoop at the foot of the rock cliff, on the northern flank of the Meteorite Valley, where only two samples had previously found. Thus, the wind scoop may be another accumulation site at FM. No samples were found elsewhere on the blue ice and moraines. However, because of the snow cover, it cannot be stated with certainty that these are unproductive areas. At the time of writing the samples are held at the Open University, the EUROMET centre for the curation, classification and distribution of meteorites to the scientific community. Hand specimen observations indicate that two or three samples might be of particular interest, possibly including one Lodranite of 4.84 g. In turn, a couple of specimens look doubtful. All the samples, apart from four specimens which remain deep-frozen because they bear evaporites on their external surfaces, have been dried and weighed. A comparison of the mass distribution of Frontier Mountain samples against all Antarctic meteorites (Fig. 2) mainly shows a deficiency of medium- to large- sized samples, namely >32 g, and a higher proportion of small-sized specimens (from 4 to 16 g), within the Frontier Mountain population. However, most of the Frontier Mountain samples are almost completely crusted and therefore it is difficult to appeal to mechanical fragmentation on Earth to account for such a distribution. Hence, the statistics suggests that some meteorites of masses ranging from circa 32 g up to 500 g are still missing at FM. Figure 1, a schematic map of Frontier Mountain (modified from [5]), appears here in the hard copy. The four meteorite concentration sites so far discovered are shown (Meteorite Valley, the wind-scoop, the accumulation site of large meteorites and the aeolian concentration area on the upwind slope of the ice depression). The general scheme for meteorite concentration at FM suggested in previous work [1-5] is also depicted by ice flow and wind directions. The ice on the plateau reaches FM from the SW and passes around it to proceed NE and ultimately feed the Rennick glacier. As the ice flows past FM some flows back towards the downstream side of the outcrop. An area of stagnant ice is formed and undergoes high ablation rates caused by katabatic winds blowing from SW. The concentration sites on the northern flank of the ice depression and in the Meteorite Valley found during previous expeditions are explained as follows: the meteorite concentration site on the upwind slope of the blue ice depression is of probable aeolian origin. Given that the dominant winds at FM blow from SW, these meteorites would have been exposed on the erosion surface of the blue ice, in an area yet to be determined but certainly between the accumulation zone and the eastern foot of FM, and then wind-blown across the blue ice field to reach the firn boundary. Meteorite Valley is an area of intense ablation of stagnant ice. The stagnation is probably due to the collision of an ice flow coming from the plateau and entering the valley's mouth from the East against a volume of ice today fed by a local glacier. Study of the Meteorite Concentration Mechanisms Active at FM: Activities and Preliminary Results Much of the field work was devoted to initiating a detailed study of the FM meteorite trap to improve the present understanding of the concentration mechanism [1, 2, 3, 4, 5] including attempts to focus on the study of the ice flow dynamics and surface aeolian mass- transport. A strain-net-network was installed in order to measure both horizontal ice flow and ablation rates. The grid was placed in the blue ice field covering areas of interest which were selected on the basis of both field evidence and a study of ice flows through the processing of LANDSAT TM images, by M. Frezzotti (ENEA, Roma). Unfortunately, due to unfavourable weather conditions, only ten out of eighteen stakes were positioned by means of the static GPS measurements. Wind directions were measured by observing wind-carved features (sastrugi and snow drifts), and its intensity by using a portable anemometer. Throughout the expedition the wind strength constantly exceeded 25-30 knots from SSW-SW. This direction is in agreement with data obtained from satellite images which confirm their prevailing character. Figure 2 (a & b), a mass distribution of (a) all Frontier Mountain meteorites (b) all Antarctic meteorites [4], appears here in the hard copy. These statistics mainly suggest that some meteorites of masses ranging from circa 30 g up to 500 g are still missing at FM. Two "rock races" were set up in the blue ice area, in order to evaluate the annual aeolian transport in relation to different rock- masses and surface morphology. These "rock-races" were positioned in areas which might yield information on the possible source regions of the localised accumulation sites of suspected aeolian origin located on the northern flank of the ice depression. Here meteorites (typically less than 170 g) were found mixed with millions of local stones arranged in banks perpendicular to the wind direction and parallel to the blue ice-firn boundary. A study was made on mass distribution of these stones, and data (Fig. 3) mainly indicates that from the base of the depression towards its upwind flank, in a NNE direction, there is a significant reduction in stone-size, from average stone-weights of 175 g, down to 35 g. This size-sorting further suggests the aeolian origin of the deposit. In addition, it seems that masses smaller than, say, 170-250 g are moved on ice surfaces by katabatic winds at Frontier Mountain. This threshold, which has to be confirmed by future controls on the rock-races is almost twice as high as that reported for the concentration site near Allan Hills [6,7] suggesting different wind's regimes. A preliminary geomorphological study of the blue ice field and of the main local moraines was also undertaken. Figure 3, a plot of the mean values of the stones sizes from different stone banks located on the upwind slope of the ice depression due W of the camp versus the distance in metres rising uphill from the bottom of the depression along wind direction (NNE), appears here in the hard copy. See text for details. Although field data are currently being processed, some preliminary observations, based on a first analysis of the aerial distribution of samples in relation to their mass, dynamics of the ice and winds, and local morphology, can be made. The general scheme for meteorite concentration at FM is that suggested in the previous works [1-5] (Fig.1), however some new elements emerge. First of all, the discovery of the accumulation site of large meteorites. Assuming that masses smaller than 170 g are moved on ice surfaces by katabatic winds at FM as discussed above, the finding of large meteorites would decrease the margin of error in identifying meteorite emergence sites for the blue-ice field. It is interesting to note that the strip of blue ice where large meteorites are found could be the source for the aeolian accumulation of small- sized meteorites found on the opposite side of the depression. In fact, since the direction of prevailing winds is 20-40 degrees toward the northeast, this area subtends the strip of wind-blown stone accumulation. Another interesting hypothesis regards the continued "productivity" of the moraine of the Meteorite Valley. Forty-one meteorites were found in this valley in 1984/85, 49 in 1990/91, and 27 this season. Assuming that this area was carefully searched, it would seem that this accumulation zone has a recharge of 8-9 meteorites per year. This would also imply that the moraine has surprisingly only been accumulating meteorites in 13-14 years! Assuming this order of magnitude is correct, the accumulation in the moraine would have formed within a time span of some decades or at least a few hundred years. This hypothesis might be in favour of a mechanism that re- exhumes a "fossil accumulation.". There is another element to be considered. A first analysis of data regarding ice flows and the morphology of FM suggest that the depression in the ice represents the collision boundary of the two ice streams which flow around FM (Fig. 1). Assuming that the aeolian accumulation of meteorites on the northern flank of the depression comes from a source to the SSW on its southern slope, the distribution of finds suggests that all the Frontier Mountain meteorites come from the southern ice stream only. Furthermore, assuming that a thorough search has been made, it seems that in the northern ice stream either do not exist conditions for meteorite emergence or, if they do, there are no suitable conditions for their accumulation. As yet, all our conclusions are tentative and require further analysis of the current data and quantitative values on ice flow vectors, ablation rates and the dynamics of aeolian transport, to be obtained through regular annual controls in the future during visits to FM. Sub-ice topography, a fundamental factor controlling the dynamics of ice, also need to be defined. Transepts are required downstream and upstream of FM, to supplement the only existing radar profile running for 6 km from the Meteorite Valley in an ENE direction, in order to have a better description of the basement's morphology and ice thickness. Acknowledgements. We wish to thank M. Frezzotti (CRE-ENEA, Roma) providing us with important data from his study of satellite images. EUROMET is financed by the EC through its Science (Twinnings and Operations) Programme; Contract No.: SCI* - CT91 - 0618(SSMA). References. [1] Delisle G.et al. (1989) Geol. Jb., E38, 483-513. [2] Delisle G. et al. (1993) Meteoritics, 28, 129-129. [3] Delisle G. et al. (1986) LPI Tech. Rpt. 86-01, 30-33. [4] Delisle G. (1993) J. Glaciol., 39, 397-408. [5] Cassidy W. et al. (1992) Meteoritics, 27, 490-525. [6] Shutt J. et al. (1986) Antarctic J. US, 21, 82-83. [7] Harvey R. P. and Cassidy W. A. (1989) Meteoritics, 24, 9-14. Reid A.* Jakes P. Zolensky M.E. Miller R. McG. Recovery of Three Ordinary Chondrites from the Namib Desert in Western Namibia Arch M. Reid, Department of Geosciences, University of Houston, Houston, Texas, USA; Petr Jakes, Department of Geology of Miineral Deposits, Charles University, Prague, Czech Republic; Michael E. Zolensky, NASA Johnson Space Center, Houston, Texas, USA; and Roy McG. Miller, National Petroleum Corporation of Namibia, Windhoek, Namibia. In 1991 we made reconnaissance searches for meteorites in selected areas of the Namib Desert in western Namibia. The 13 meteorites that have been described from Namibia include the very large Hoba and Gibeon irons, and four chondrites (Gobabeb, Namib Desert, St. Francis Bay and Witsand Farm) that were found in the region of the western desert. To our knowledge the area had not been visited previously with the express purpose of looking for meteorites. Three new ordinary chondrites were recovered as a result of this search. The Namib is a long narrow desert region in western Namibia, extending approximately 2000 km from the Olifants River in northwestern South Africa to the Carunjamba River in southern Angola. While the determination of the age of desert surfaces is somewhat controversial, there are significant regions in western Namibia with surface ages believed to be at least 5 Ma (1). Aridity is high throughout the area; however, there is moisture along the Atlantic coastline and immediate interior, due to the common presence of coastal fog banks. The area south of the Kuiseb River includes some of the world's largest dunes. We made preliminary searches of five different regions in western Namibia. 1. The fan delta and older terraces along the course of the Omaruru River, to the north and east of Henties Bay. Despite the fact that there are deflation surfaces, with few coarse rock fragments, on these older terraces, no meteorite material was encountered. 2. The older river terraces south of the Swakop River, to the east of Swakopmund. These surfaces are similar to area 1 and no meteorite material was found. 3. Deflation surfaces in the broadly flat-lying region to the east of Walvis Bay, close to the Namibia-South Africa border. In this region three fairly well-preserved ordinary chondrites were recovered. 4. Within the Namib Sand Sea, on deflation surfaces in inter-dune corridors, between the major longitudinal dunes. The area we examined, lying to the west of Tsondabvlei and south of Gobabeb, is fairly typical of the central desert region, with extremely elongate high dunes alternating with flat inter-dune corridors. One chondrite, described by Fudali and Noonan (7), has previously been found in this area. While the flat corridors are apparently fairly stable features, we could find no meteorite material within them. 5. The area in and around the Roter Kamm impact crater in the extreme southwest of Namibia, in a remote desert region to the north and west of Rosh Pinah. In examining the crater and collecting crater-related samples, we took the opportunity to search the area for meteorite material, but without success. Our search method began with the selection of the above regions as having the best potential, and continued with the selection of sub-areas based on examination of topographic maps, and selection in the field of specific search areas, based on the presence of deflation features and the absence of larger rock fragments and grass cover. Searching was purelv visual with 3-4 searchers in parallel or random search patterns aimed at examination of all rock fragments greater than a few cm diameter, within the selected area. All five of these areas are prospective collecting sites because of the existence of deflation surfaces of considerable age, in a region noted for its extremely arid climate. Reasons for the lack of recovery at four of the sites are related to proximity to the Atlantic coast and the effect of the coastal fogs, to sand movement and the consequent burial of old land surfaces, to our lack of success in recognizing the oldest land surfaces, to the difficulty of recognition of meteorite material (some recovered samples show a surface varnish due to prolonged desert exposure, with embedded quartz grains), and to luck. The region in which we did recover meteorite fragments is east of Walvis Bay, just within the Namibian border, on a series of deflation surfaces at approximately 23 degrees 5.0' S, 14 degrees 42.9' E. Each of the meteorites was almost wholly exposed at the surface. Two of the samples were single stones whereas the third comprised 27 fragments in a area of approximately 2 square meters. The samples are all fairly close to the border station of Rooikop, which mav provide an appropriate name for the meteorites. The largest of the three meteorites is a single severely weathered stone with thin pervasive iron oxide veinlets, that weighed 1.039 kg. The sample is an H-group chondrite with constant composition olivine, Fo80.9, orthopyroxene Wo1.3En8l.9Fsl6.8, clinopyroxene Wo47.1En46.7Fs6.2, and plagioclase Or6.3Ab80.5An13.2. Minor chromite and phosphate occur along with troilite and metal. Chondrules are generally poorly defined and the meteorite is classified as an H6 (3). The least weathered of the three meteorites is a single stone that retains a fusion crust and weighed 0.902 Kg. The interior surface has prominent chondrules up to 2 mm diameter and the meteorite is an L-group (L4) chondrite with homogeneous olivine, Fo76.6, and orthopyroxene, Wol.3En77.9Fs20.8. Some multiply twinned clinobronzite is present, along with fine-grained devitrified glass, and minor chromite, troilite and metal. The third meteorite comprises 27 small fragments with the largest pieces weighing 0.401 and 0.362 kg. It shows well-developed chondrules up to 1 mm diameter, but is weathered with many fine oxide veins. It is also an L-group chondrite (L5) with homogeneous olivine, Fo75.4, orthopyroxene, Wo1.2En76.3Fs22.4, and clinopyroxene, Wo45.3En46.2Fs8.5. Chromite, troilite and metal are present in minor amounts and there is one occurrence of high silica glass. The recovery of three new chondrites is encouraging, considering the vast area for potential meteorite recovery. We are currently trying to organize a follow-up visit to expand the search within the area east of Walvis Bay, and also to explore new search areas. References: (1) J. Ward and F. Corbett (1990) in Namib Ecology; 25 years of Namib Research. Transvaal Museum Mono. 7, 17-26. (2) R. F. Fudali and A. F. Noonan (1975) Meteoritics 10, 31-39. (3) A. M. Reid, P. Jakes, M. E Zolensky, and R. McG. Miller (1992) Abstract, Lunar Planet. Sci. Conf. 23, 1135-1136. Franchi I. A.* Delisle G. Jull A. J. T. Hutchison R. Pillinger C. T. An Evaluation of the Meteorite Potential of the Jiddat Al Harasis and the Rub Al Khali Regions of Southern Arabia Over the years there has been considerable success from organised meteorite search programs in hot desert areas of the world, such as the Nullarbor Plain in Western Australia (Bevan and Binns, 1989), Roosevelt Co., in New Mexico (Huss and Wilson, 1973) and Reg el Acfer and Tanezrouft in Algeria (Bischoff and Geiger, 1992). However, the discovery of new areas is important if our meteorite collections are to continue to grow as the rate of return from these existing areas diminishes or access becomes problematic or commercial exploitation becomes dominant. Therefore, this paper is an evaluation of the potential of areas in the south-eastern part of the Arabian peninsula for any meteorite recovery program. Two areas stand out as having yielded relatively high concentrations of meteorites: Jiddat al Harasis, a large sand and gravel plain in central/southern Oman and the Rub al Khali, a huge sand sea in southern Saudi Arabia. Both areas are extremely sparsely populated yet 24 meteorites have previously been found in this region (Fig 1). Indeed, the observation of relatively large numbers of meteorites was first commented upon by Holm (1962). Five of the meteorites were found in the early 1930s during the first crossings of the area by Europeans and almost all of the remainder were recovered after opening up of the region to oil prospecting operations in the 1950s and 60s. The Arabian peninsula has generally been arid since the start of the Pleistocene and hyper-arid since about 17000 BP (McClure, 1978). On the basis of the meteorite concentrations, the arid climate and the sparse human population a six day reconnaissance of the Jiddat al Harasis by the first two authors of this paper was conducted as part of the EUROMET program in October 1993. During this reconnaissance a further two samples were recovered. The Jiddat al Hirasis is an area of about 60000 km^2 with an average altitude of about 200m above sea level (Fig 1). The terrain is predominantly composed of Oligocene/Miocene grey/buff limestones and marls with some chalks. The sediments are almost flat lying with a very gentle dip of about 6 degrees to the north. The effect of this northward dip is to create an internal drainage pattern resulting in a number of playa deposits across the region. The total variation in relief is no more than a few tens of metres with very gentle inclines, although occasionally small broken scarps have developed. The surface of the plain is variable in nature, ranging from quite blocky limestone rubble (angular fragments upto 15cm) to coarse sand/gravel soils to chalky sands and ranging in colour from grey through brown to white. The number of dark stones present on the surface, a key factor in determining the probability of identifying a meteorite, was generally fairly low. There is no apparent major sediment input into the plain, although wind sometimes blows from the north and transports sand from the Rub al Khali. Many of the surfaces appeared to be ablation surfaces, either stripped of soil or with concentrations of stones on the soil surface. Vegetation ranged from sparse grasses and occasional small trees in the south east of the region to essentially zero in the west and the north (with the exception of wadis). During the six days in the field eight different localities were searched, usually by foot, sampling a range of different terrains all the way across the region (Fig 2). Logistics in this region were very straightforward due to the Muscat-Salalah highway running through the north of the region with a number of well maintained (but incompletely mapped) graded roads servicing the various oil production facilities. The two meteorites found during the reconnaissance trip were 1.75kg and 0.4g and have been tentatively classified as H4 and L7 respectively. The larger sample is quite weathered and was recovered as 20 fragments spread over ~2m. The similar degree of weathering and the proximity to one of the known Jiddat al Harasis stones (Fig 2) strongly suggest that this may infact be part of the same fall. Due to the small size of the second meteorite and the lack of positional information about the Hajmah finds it is difficult to determine whether this sample is also paired. Overall this area showed considerable promise as a possible meteorite search area, a further test of its fruitfulness will be conducted with a larger party with more concentrated searching. However, the intense oil exploration and production activity over the past 40 years has left much of the ground thoroughly covered in tyre tracks, perhaps suggesting that some meteorites in the area may have already been collected. As an example, within a 20m radius of the large meteorite there were 9 sets of tyre tracks, perhaps indicating that most of the vehicles were either driving too fast to spot meteorites or were not interested in small dark stones. The Rub al Khali is the largest sand sea (~550000km^2) on the planet and currently one of the most arid. Although initially such an area would not normally be considered as a potential meteorite bearing region the number of meteorites relative to the total number of people in, or who have crossed the area suggests that more meteorites exist there. The question - is how many more? Many of the meteorites which have been recovered from this area have been rather badly weathered, suggesting either long terrestrial residence times or poor preservation conditions. At the eastern end of the Rub al Khali is a large playa deposit (~3500km^2) with extensions in inter-dune areas extending up to 200km further west (Fig 1). Although this area has been hyper-arid or arid for most of the pleistocene there have been some periods of more humid conditions (McClure, 1978), during which there would have been development of the playas. Obviously, the presence of saline water, even for short periods, could greatly accelerate the weathering of any meteorites. However, a preliminary survey of terrestrial ages on the Rub al Kahli and Jiddat al Harasis meteorites (Table 1) display a range of ages from 6,400 to 31,100 years. This range is comparable to other desert meteorite collections (Jull et al, 1993), and although the meteorites from this area probably are somewhat more weathered this does not appear to have significantly affected their probability of preservation. Table 1. 14C terrestrial ages (method as in Jull et al, 1989) of meteorites from Jiddat al Harasis and Rub al Kahli. Jiddat al Harasis Hajmah (a) Ureilite 18,300 +- 1,700 Hajmah (c) L5-6 15,200 +- 1,300 Tarfa L-6 15,200 +-1,300 Jiddat al Harasis H4 31,100 +- 2,300 Rub al Khali Suwahib (Adraj) L4 6,400 +- 1,300 Suwahib (Ain Salah) H6 27,500 +- 1,600 Satellite images show that there are large tracks of inter-dune areas (several kilometres wide by tens of kilometres long), particularly in the southern parts of the Rub al Khali, where it would be straightforward to conduct systematic searches. It would also be interesting to establish whether the very large, and relatively stable dunes in this region have produced any local meteorite concentrations. However, due to the remoteness of this area and the difficulty of crossing sandy terrain there are many logistical and some political problems associated with attempting to operate in this region. References.A.W.R. Bevan and R.A. Binns (1989) Meteoritics 24, 127-133. A. Bischoff and T. Gieger (1992) Met. Bull. 73, Meteoritics 27, 477- 478.A.J.T. Jull, D.J. Donahue and T.W. Linick (1989) Geochimica et Cosmochimica Acta 53, 2095-2100.A.J.T. Jull, F. Wlotzka, A.W.R. Bevan, S.T. Brown and D.J. Donahue (1993) Meteoritics 28, 376-377.D.A. Holm (1962) Amer. Jour. Sci. 260, 303-309. G.I. Huss and I.E. Wilson (1973) Meteoritics 8, 287-290.H.A. McClure (1978) In: Quaternary Period in Saudi Arabia. Springer-Verlag, 252-263.EUROMET is indebted to Dr. Hilal Al Azri at the Minestry of Petroleum and Mines and Mr Ralph Daly at the Office of the Advisor for Conservation of the Environment, in Muscat. We are also very grateful for the help of Mr. Nigel Winser at the RGS in setting up all the necessary conections. EUROMET is is supported by the EC through its Science (Twinnings and Operations) Programme; Contract No.: SCI*-CT91-0618(SSMA). Figure 1, a map of the SE portion of the Arabian peninsula showing all meteorite finds in the Rub al Khali and Jiddat al Hirasis, appears here in the hard copy. Also shown is the extent of the large playa and the larger area of playa exposed in inter-dune areas. Figure 2, a map of meteorite locations and search areas on the Jiddat al Harasis, appears here in the hard copy. Meteorite locations shown with shaded circles (larger circles show position of new finds). Search areas are shown with a large S. The locations of the three meteorites round Hajmah are not accurately known - their given positions are those of the oil camp where they were first identified as meteorites. Similarly, the most westerly meteorite is also poorly referenced. Zolensky M. E.* Schutt J. W. Reid A. M. Jakes P. Martinez de los Rios E. Miller R. M. Locating New Meteorite Recovery Areas Introduction: Several years ago two of us (JWS and MEZ) resolved to visit Roosevelt County, New Mexico, to look for meteorites. We then decided that rather than search where others had before us, we would attempt to find a new meteorite concentration surface. Accordingly, we chose the general area south of Roosevelt County as the exploration target, and thought through the exercise of "how to predict where meteorites can be found". JWS had considerable experience in the Antarctic Search for Meteorites, having served as ice guide and all- around expedition expert for many field seasons with Bill Cassidy. From the Antarctic Experience and by extension from the Roosevelt County finds we knew that target areas should exhibit the following characteristics: (1) Low Humidity: Aside from the polar deserts, most deserts straddle either the Tropic of Cancer or Capricorn (23.5 degrees N and S latitude, respectively). These deserts include the Sahara, Arabian, Victoria (Australia), Kalahari and Namib (S. Africa), Sonoran (N. America), Atacama (S. America), Registan (Afghanistan), Baluchistan (Pakistan), and Great Indian Desert. The Takla Makan (north of Tibet), Gobi and Kashmir (W. China) Deserts and those in the western United States are examples of topographic deserts, located deep inside of continental masses, cut off from oceanic bodies and humid winds by mountains. Any of these deserts are potential target areas, few have been the subject of systematic searches. These deserts are all located in Figure 1. (2) Minimal Fluvial Activity: Stream activity should be minimal, so that little terrigenous material has been deposited or removed during the lifetime of meteorite accumulation and excavation. (3) Minimal Input of Terrestrial Rocks: Anyone who has spent time in Antarctica searching for meteorites in a glacial moraine crowded with black rocks will recognize the desirability of this characteristic. Surfaces are far more easily searched if there is little terrestrial material to obscure the meteorites. This requirement can be satisfied by geologically quiet, non fluviatile areas with subdued topography. (4) Rapid Burial of Meteorites: For Antarctic meteorite concentrations, it is generally thought that most meteorites fall onto the icecap far from the eventual stranding areas. They are subsequently buried by snow which gradually compacts through firn to ice. In other places on Earth the most efficient burial may be performed by blowing sand. In any case, meteorite preservation is facilitated by rapid burial, and removal from the atmosphere. (5) Recent Deflation: In Antarctica the meteorites are concentrated by a combination of glacial movement and ablation of the ice carrying the meteorites. Elsewhere, the corollary to (4) is relatively rapid meteorite excavation, by wind (deflation) or burrowing mammals. In the S.W. United States deflation is a cyclical phenomenon; where wet periods alternate with dry periods. During the wet periods calcite, the principle cementing agent for local coversands, can be leached away, promoting the removal of loosened sand grains during dry periods. Mammals can contribute to this process; in the 1950's large mammals working for the U.S. Department of Agriculture recommended to farmers living in Roosevelt County (New Mexico) that they plow up large tracts of grassland. Unfortunately, this conservation effort immediately preceded a period of drought and substantial winds, during which the previously fertile soils of the grasslands were stripped away leaving large deflation basins in their place. The local economy was damaged, but hundreds of meteorites were exposed and subsequently recovered. A corollary of this criterion is that there must be minimal ground cover by plants. (6) Degree of Induration of the Stranding Surface: The final stranding surface must be sufficiently indurated to support meteorites so that they may be recovered. Even a thin layer of loose dust or sand, blanketing the ground, can effectively disguise meteorites in the same way that a recent snowfall can in the Antarctic. In the southwestern U.S. Pleistocene lacustrine beds make good stranding surfaces, as they are rich in Ca-carbonates and consequently are well-cemented. Using these criteria, we should be able to select specific regions of Earth's surface for future meteorite recovery expeditions. How well will this scheme work? Lets examine three subsequent expeditions which were undertaken with these criteria as a guide. Lea County, New Mexico: The high numbers of meteorites recovered in Roosevelt County indicated that additional finds might be made in similar landscapes to the south, in neighboring Lea County. In 1988 JWS and MEZ made foot searches of recent deflation basins exposing Late Pleistocene eolian sediments and topsoil in the region immediately south of Jal, New Mexico [1]. These deflation basins appeared to satisfy all criteria save number 4. The deflation basins in the Jal vicinity do not generally exceed 300 m by 50 m in size, and the floors are a well- indurated, red to gray colored calcareous sandstone of possible lacustrine origin. These floor sediments appear to be 50-100 ka in age, in analogy to the deflation basins in Roosevelt County [2]. We found two chondrites in our foot searches (one H and one ungrouped), hardly a high concentration, but suggestive of further discoveries in the future. We are now performing a survey of additional similar sites immediately to the east in Texas. The widespread nature of Pleistocene lacustrine deposits in the Desert High Plains of the U.S. suggests that additional meteorite concentration surfaces will be found. It is probable that the rate-controlling step for these discoveries will be sudden intense deflation episodes. Atacama Desert, Chile: The Atacama Desert is one of the driest regions on Earth. In 1991 EMR and MEZ made jeep and foot searches of large deflation surfaces situated between the Chilean city of Antofagasta and the seaside fishing village of Mejillones (an area called "Pampa"). The deflation surfaces are immediately inland of a broad coastal mountain (Cerro Moreno) which insulates the region from oceanic moisture. Pampa is an elevated beach deposit, and the surface is poorly indurated (so meteorite fragments are easily disguised by dust). Considering the remaining uncertainties concerning the local geological history, only criteria 1 and 2 are clearly satisfied by the Pampa area. Nevertheless, between those samples found previously by Edmundo Martinez and those found on the 1991 expedition, we (in concert with Rene Martinez) have described five (different) new L chondrites [3]. There appears to be significant future potential at this and adjacent sites. It is interesting that only L chondrites have been found at Pampa. One might think that the lower amount of metal in L chondrites, relative to H chondrites, might stabilize them with respect to terrestrial weathering. However, there is apparently no real indication of an enhancement like this for, say, the Antarctic finds relative to worldwide falls (Michael E. Lipschutz, personal communication, 1994). Namib Desert, Namibia: The Namib is one of the wettest deserts that have been searched for meteorites. It is located along the south-western coast of Africa, with little in the way of barrier topography to shield it from oceanic moisture, which appears here in the form of frequent morning fog. The age of the Namib surfaces is subject to controversy, but is generally believed to be at least 5 Ma in places (see [4]). In 1991, AMR, PJ and MEZ performed reconnaissance foot searches of four different areas in and adjacent to the Namib, selected in advance with the assistance of Justin Wilkinson. These target areas were (1) along the fan delta of the Omaruru River, north and East of Henties Bay, (2) along river terraces south of the Swakop River, east of Swakopmund, (3) on deflation surfaces and inter-dune corridors within the Namib Sand Sea, west of Tsondabvlei and south of Gobabeb, and (4) on deflation surfaces east of Walvis Bay. Meteorites were only found in the last mentioned locale; these being three ordinary chondrites (two unpaired L's and one H) [4]. "Recent" fluvial activity may account for the lack of significant meteorite concentrations in locales 1 and 2. It is more difficult to account for the failure to locate meteorites in locale 3, where deflation surfaces of tremendous size were present (measuring up to several kilometers across each). We noticed that the surface at locale 3 was not well indurated, being covered with fine sand and lag deposit to a depth of several cm, and this may account for the apparent lack of meteorites, i.e. they were shielded from view by a thin layer of sand. The meteorite-producing locale, number 4, satisfied criteria 2, 3 and 5. It was also sufficiently inland to be fairly dry (criteria 1). The surface at this site was not particularly well indurated, however the sand blanket was no more than 1 cm thick, in contrast to the other sites we searched in Namibia. We conclude that, in addition to the 6 criteria proposed above, two other criteria should be considered when evaluating potential new areas for meteorite exploration: (7) Sand Cover Thickness: If the surface is not well-indurated then the depth of any loose sand cover is a critical factor; this should be very thin (<=1 cm). (8) Age: The age of the meteorite accumulation surface should be at least 100 ka, to permit accumulation of a reasonable number of meteorites. This rather arbitrary age is proposed by analogy to the Roosevelt County meteorite accumulation surfaces. We emphasize that luck plays a definite role in meteorite exploration, as does the very important ability to differentiate very weathered meteorites from the local rock population. References: [1] Zolensky M.E. et al. (1989) Meteoritics, 24, 227-232; [2] Zolensky M.E. et al. (1992) Meteoritics, 27, 460-462; [3] Martinez R. et al. (1992) Meteoritics, 27, 254-255; [4] Reid A.M. et al. (1992) Lunar and Planetary Science XXIII, 1135-1136. Gerel O.* Bischoff A. Schultz L. Schluter J. Baljinnyam L. Borchuluun D. Byambaa C. Garamjav D. The 1993 EUROMET/Mongolian Expedition to the Gobi Desert: Search for Meteorites The Gobi desert in central Asia belongs to the largest deserts on Earth. It is about 2000 (W-E) x 1000 (N-S) km in size and is located in southern Mongolia and northern China. Based on scientific contacts between the University of Hamburg and the Mongolian Academy of Sciences a plan for a EUROMET meteorite search expedition together with the Academy of Sciences of Mongolia was worked out in 1992. The expedition started on August 28, 1993 in Ulaanbaatar and took place between August 28 and September 13. The route of the expedition is given Fig. 1: Ulaanbaatar-Mandal Gobi- Tugalin Bulen-Dalanzadgat-Chinese Border-Tabun Khara Obo-Sainshand- Airag-Ulaanbaatar. The central part of the Gobi desert is located south of Dalanzadgat and belongs to the variscian South-Mongolian Zone. Characteristic for this zone is the abundance of rocks derived from basic volcanic explosions in the silurian and devonian period in combination with deep-sea cherts and rocks of the ophiolite association. The platform cover from upper cretaceous to cenozoic consists of coarse- grained molassic red-coloured sediments, effusiv basaltic rocks, and gravels and sands. The expedition spent most of the time in these areas searching for locations with light-coloured surface rocks without dark volcanic constituents. Since geologic mapping for most areas exists only on a 1:1000000 scale, it was quite impossible to find appropriate areas for meteorite search. A small aeroplane was chartered to get some more information about the Mongolian part of the Gobi desert. Indeed, some areas were found that appeared to be interesting for search activities (Borzougiin and Galbien Gobi). For several days, such areas were scanned by seven people on foot; in addition some search was performed by car; but no meteorite could be found. Why meteorites were not found? The expedition in the Gobi desert was faced with one severe problem. The summer 1993 was one of the wettest in the Gobi desert for the last 50 years. Thus, it was not possible to reach all areas of interest, because of too much water in some normally small rivers that could not be crossed. Also, the desert was "green" with lots of vegetation. This fact led to another problem. With the vegetation also thousands of antelopes and khulans (kind of wild horses) were moved south (into the normally dry area) causing black "fall-out" everywhere. On the other hand as stated above, in many locations of the Mongolian desert volcanic rocks occur. The occurrence of black volcanic rocks within the sediments makes the recognition of meteorites almost impossible. We only found small areas without these black rocks from volcanic activity. Based on many dried water streams in the Gobi desert we believe that once in a while heavy rain is falling changing the morphology of the surface and reworking the surface sediments. In addition, typical heavy winds take part to modify the upper meters of the sediments. Thus, there may be large problems to find very old unprocessed areas. Tabun-Khara-Obo Crater The expedition passed the Tabun-Khara-Obo crater, which is located in southeastern Mongolia about 470 km SSE of Ulaanbaatar and 95 km SSW of Sainshand. Based on Landsat I photographs McHone and Dietz [1] described the impact crater as an 1.3 km (diameter) sand-filled crater with a high degree of circularity. It is suggested that the crater is about one million years old. The elevation of the wall crest above the bottom of the depression is on average 20-30 m and reaches a maximum in the east of 50 m [2]. Microschists and diorites within the wall are intensely deformed. Breccias occur in form of lenses mainly within the inner crater wall and are several meters thick. A block of fine-grained, fragment-rich impact melt was found inside the crater. Some breccias were also found outside the crater up to about 500 m from the crater rim. Positions of the highest elevations (4 positions) of the crater rim were taken by GPS. Based on these measurements the crater is certainly larger than 1.3 km in diameter and probably close to 1.8 km in diameter (Fig. 2). A profile of samples was taken from Tabun-Khara-Obo crater. At 11 positions about 60 samples of the representative rocks were collected (Fig. 2). A study of these rocks is in progress. Acknowledgements: This study and the expedition was supported in part by the DAAD (Deutscher Akademischer Austauschdienst), the Mongolian Academy of Sciences, and by the European Economic Community SCIENCE (Twinning and operation) programme, contract No. SC1*-CT 91-0618 (SSMA). We thank Dr. T. Boldsukh (General Scientific Secretary of the Mongolian Academy of Sciences) for his generous support and the drivers of the jeeps during the field work. [1] McHone and Dietz (1976) Meteoritics, 11, 332. [2] Masaitis et al. (1980) Geology of Astroblemes. Fig. 1: Route of the reconnaissance meteorite search in Mongolia. Positions taken by a GPS instrument are given by small dots; indicated are a few villages as well as the impact crater Tabun-Khara-Obo. Triangles are campsides. Fig. 2: The Tabun-Khara-Obo crater in Mongolia. Wednesday, July 20, 1994 MECHANISMS AND CHALLENGES 4:15 - 5:15 p.m. Chair(s): J. O. Annexstad Delisle G.* Storage of Meteorites in Antarctic Ice During Glacial and Interglacial Stages Introduction: Antarctic meteorites are better preserved over longer time periods than material from hot deserts. Ice is clearly an ideal host material for storage. Nevertheless, there is evidence that many meteorites reach periodically the blue ice surface, where they moisten and refreeze and experience similar weathering and decay processes as their counterparts in hot deserts, before they are being reburied again and isolated in deeper ice levels. The purpose of this paper is to discuss two points: (1) The field situation in which Antarctic meteorites are found today is most likely not typical for their long time storage mode. (2) There is field evidence for movement of fluids and/or gaseous components in surficial ice during the Antarctic summer season. Antarctic meteorite traps in general: The major Antarctic meteorite concentrations are found at elevations between 1900-2400 m above sea level. Computer simulations on the response of the East Antarctic ice sheet to glacial stages (climatic deterioration) predict for this "meteorite trap terrain" a higher ice stand and shallower surface slope than today [1-3]. The reasons are as follows: the decrease in precipitation during a glacial stage (cooler and dryer air) over central Antarctica in combination with the cooling of the ice cap causes a thinning of the central East Antarctic ice sheet and thickening along the coastal regions. However, during an interglacial stage, the process reverses (coastal ice shrinks, the central ice sheet thickens). A computer model by [3] postulates that the area of zero ice stand fluctuation is located near the 2700 m elevation line. The known meteorite traps are all located on the coastal side of this "hinge line." They should experience ice thinning (e.g., 100-300 m according to [3]) during an interglacial stage and a return to thicker ice during glacials. The duration of "glacial stages" is on the order of 110 b.y., while interglacial stages (as we currently live in) last for roughly 15 b.y. What are the consequences during glacial stages? The shallower surface slope and the cooling of the ice tend to reduce ice velocities and mass transport. An increase in ice thickness in combination with an unchanging ice ablation rate (here assumed for the sake of the argument) tend to speed up ice movement. What is the net effect? The case of the Allan Hills: The on average 200 m thick, westward sloping blue ice along the W-flank of the Allan Hills is actively decaying today [3]. The once thicker blue ice was either previously transported to the site from the Antarctic interior or was produced locally during a period of a much higher ice stand sufficient to exert the required overburden pressure to form this dense type of ice. Latter scenario is less likely. A growing snow and ice cover would readily have moved across the shallow Allan Hills towards the coast, for which we have no field evidence (no lateral moraines evident in the field). Above scenarios preclude the ice depression east of the Allan Hills Main Icefield escarpment to have existed at the same time. A simplified computer simulation of the ice flow demonstrates that the ice ablation rate during glacial stages must have been much lower than today or even absent, which is equivalent of saying that a large portion of current blue ice fields was then covered by firn and snow. Considered is a 500-m-high and 8-km-long vertical segment of blue ice. The ice thickness on the left boundary is kept constant (representing the fast moving ice stream between the Main and Near Western Icefield [4,5], not subject to ice sublimation). The upper surface of the ice is exposed to ice sublimation at a rate of 4 cm/a for t > 0. Eventually, a balance will develop between the ice loss by sublimation and the advected ice. The computer simulation creates an ice surface slope (which is the force driving the advection of ice) of 130 m over a distance of 8 km , which is almost exactly the situation observed in the field today (line a in drawing). If we lower the assumed ice temperature in the model by 10 degrees, the surface slope increases to 190 m over the same distance (line b). The reason is: cooler ice is stiffer and needs a greater surface slope (greater driving force) to maintain the amount of advected ice necessary to balance the ice loss by sublimation. None of these models would account for the presence of blue ice along the western flank of the Allan Hills. In the past, the local ice sublimation rate must then have been much lower or the regional ice level higher or both. This scenario was tested by the model now assuming a 600 m high vertical segment with a length of 12 km--equivalent to the distance between the Allan Hills and the ice stream. The ice sublimation rate was reduced to 2 cm/a. The calculated surface slope intersects the western flank of the Allan Hills about 50 m below the hilltops (line c in drawing). It should be remembered that the model assumes constant ice thickness. The rising bedrock near the Allan Hills would in reality steepen the surface slope even further above it. The sublimation rate would have to be lowered further, if the existing bedrock topography would be incorporated. Figure 1, showing the approximate outline of the surface and subice topography along a from the Allan Hills westwards and (2) the theoretical ice surface slope for (a) an ice temperature of -30 degrees and an average sublimation rate of 4 cm/a; (b) an ice temperature of -40 degrees and an average sublimation rate of 4 cm/a; (c) an ice temperature of -40 degrees and an average sublimation rate of 2 cm/a, appears here in the hard copy. From this consideration it is concluded that field observations strongly suggest in connection with the above outlined computer simulation on the ice flow under climatic conditions of the last glacial stage an higher ice level and a much reduced sublimation rate in the past. The case of the Frontier Mountain: As discussed previously (see, i.e., Fig. 20 in [6]), the current surface topography at Frontier Mountain most likely evolved through various stages starting from a situation with a regionally higher ice stand and a much reduced surface ablation rate. Consequence: The postulated processes reduce the exposure rate of meteorites during glacial stages to a minimum. Meteorite occurences would then be covered by a substantial layer of firn and ice. This scenarion pictures the meteorite traps to operate essentially only during interglacial periods (as today). The storage environment in ice today: Several boreholes were drilled into ice during the German GANOVEX -VII expedition to Victoria Land (1992/93). The drilling device (a detailed description can be found in [7]) was an electrically heated sonde that melted its way through the ice. The meltwater was removed from the hole by a bailer. One borehole was drilled into blue ice at a location (75 degrees 54,182S; 158 degrees 32,778E) due north of Ambalada Peak. No meteorite had ever been found on this blue ice. The site nevertheless offers the typical conditions of Antarctic blue ice fields. The monitoring of the drilling operation showed that the recoverable amount of meltwater from the borehole varied. The theoretically predicted amount of 5.7 l per m of borehole was recovered down to a depth of -4 m implying an impermeable ice wall. A sharp reduction down to less than 1 l per m occurred below that depth down to bottom at -10 m. These observations are interpreted as follows: The blue ice had developed under tension numerous open fissures, which are not obvious at the ice surface. During the summer season any developing fissure is resealed rapidly (water vapour or meltwater?). This process is apparently active down to a depth of about -4 m. Alternatively thermal expansion of surficial ice in response to the advancing thermal summer wave closes open cracks. Fissures below this level are not affected and remain open. The permeability of the ice below this level is apparently sufficient to drain about 30 l of water within half a day from a borehole with a diameter of 8 cm. References: [1] Oerlemans J. and Van der VeenC. J., eds., (1984) Reidel, 217 pp. [2] Hybrechts P. (1990) Ann. Glaciol., 14, 115-119. [3] Delisle G. (1993) J. Glaciol., 39, 397-408. [4] Schultz L. et al. (1990) Ant. J. US, 94-95. [5] Delisle G. and Sievers J. (1991) JGR, 96, 15.577-15.587. [6] Cassidy W. A. et al. (1992) Meteoritics, 27, 490-525. [7] Zeibig M. and Delisle G. (1994), Polarforschung (in press). Harvey R. P.* Moving Targets: The Effect of Supply, Wind Movement and Search Losses on Antarctic Meteorite Size Distributions The size distributions of Antarctic meteorite collections are influenced by many distinct factors, including the original supply of meteorites from space, losses to weathering and wind movement, and searching techniques. Many of these factors have been modelled, both empirically and theoretically, by previous researchers, usually with regard to geological materials other than meteorites. In this work I will combine new and established models for application to studies of the size distribution for all Antarctic meteorites (AAM) from [1]. The resulting model is dependent on measurable variables and has application to meteorites from other collection sites as well. Supply: Accepted models for the supply of meteorites to the Earth's surface usually take the form of a power-law, based on empirical studies of impact distribution on orbiting spacecraft and studies of mechanical breakage related to impact phenomena [2,3]. The basic premise is that fragmentation events produce an exponential increase in specimens, each of which is exponentially smaller. Power-law distributions consequently take the form of a straight line on a log-log plot of mass (size) vs cumulative number of specimens (Fig. 1b). As a result, at diminutive sizes the number of small particles can be dramatically large. However, these very small particles are often more susceptible to loss phenomena such as removal by wind, further breakage, and search inefficiencies, as discussed below. Past researchers have considered this when fitting power-laws to observed meteorite size distributions at the earth's surface, constructing their models from the very largest fragments- the majority of the mass but a minority in number [4-9]. Unfortunately these power-law models then become extremely sensitive to the slope of a line modelled from very few accurately known data points [1, 10, 11]. Figure 1. Size distributions for the AAM dataset (n= 5715) a. shows the size distribution plotted with a non-logarithmic count of specimens along the vertical axis, and a series of mass bins, doubling in size to the right, on the horizontal axis. b is the same dataset plotted with a cumulate-log vertical scale. The power-law shown was fitted to all meteorites larger than 64 g and an x-intercept equal to the size of the largest AAM meteorite (407 kg). Weathering: Fragmentation due to weathering and abrasion processes such as frost wedging, salt production, wind abrasion, chemical exsolution, and biological actions [12] also contribute to the number of specimens ultimately found on a stranding surface. The vast majority of meteorites from Antarctica exhibit significant weathering features, even those found still enclosed in ice [13-17], due to exposure to saltating snow and ice particles, long duration freeze-thaw cycles, and evaporite formation [15, 18-20]. Most empirical fragmentation models produce bell- shaped size distributions with a maximum where a rock has been reduced to a collection of resistant mineral fragments. While such distributions show a good empirical fit to the AAM dataset, they do not incorporate the initial power-law supply commonly accepted, and controlling variables often are not physically measurable. Theoretical models of fragmentation, however, are easily incorporated into a modelled power-law supply, by increasing its slope. The model presented here defines the combined supply and fragmentation power-law with two variables. The X-intercept is the mass of the largest observable meteorite: for Antarctica this is roughly 400 kg. The power-law slope defines how many meteorites of sizes smaller than the X-intercept are found; literature values vary between 0.7 and 1.2 [8]. Figure 1 shows a typical power-law fit to the AAM dataset as used to formulate the combined model. Wind loss: Like weathering, wind movement of particles has been extensively modelled under conditions appropriate to desert and polar environments (see [21-24] for summaries). These models, both empirical and theoretical, attempt to formulate the wind velocity necessary to initiate particle movement given various conditions of air density, particle size, adhesion and frictional forces, and surface conditions. Although these models can become quite convoluted, Antarctic meteorites present a very tractable problem by virtue of their resemblance to an ideal physical model. Meteorites generally are well exposed to the wind, are roughly spherical bodies lying on a relatively frictionless, level plain, and complicating factors such as the vertical wind profile, sun- cupping, pedestal and wind-scoop formation are easily incorporated into the formulation of the relation between meteorite size and the threshold windspeeds necessary to establish movement. Annual wind velocity distributions are available for several locations on the East Antarctic plateau [25]. These wind velocity distributions take two general forms; an exponential form, where high winds are relatively rare but milder winds are omnipresent; and a gaussian form, where an intermediate windspeed is most common (Fig. 2a) Knowing the probability of a windspeed's occurrence during a given year and calculating the size of meteorite that wind can move yield a relatively simple relationship between specimen size and the probability of particle movement. As expected, the probability that a particle will move is strongly anti- correlated with size. Larger particles have a low probability of threshold movement, and thus while they may occasionally be in motion it should be a relatively rare event. Small particles have very high probabilities of encountering threshold wind velocities, and may effectively be in constant motion, greatly increasing their likelihood of leaving meteorite search locations. The model presented here calculates the probability of particle movement based on a choice of either exponential or gaussian windspeed distribution, the mean and variance of windspeed, and threshold velocity variables as discussed above. The number of meteorites within each bin of the size distribution is then reduced by the probability those meteorites would move during a single year. This assumes that particles with 100% probability of movement during a single year are certainly lost over a longer time period. Figure 2. Probability of threshold velocity occurrence in Antarctica. a. shows measured wind velocities at two sites along the Transantarctic Mountains during 1992. Pt is the percent of time winds of various velocities were recorded. b. shows exponential and gaussian curves fit to the data from a, plotted as probability of threshold velocity occurrence (Pm) in a given year vs mass. Note that although winds of higher speeds occur in the exponential regime, particles with masses <200 g have a higher probability of moving in a gaussian regime. Searching efficiency: As in the case of wind losses above, meteorite collection presents a very tractable modelling problem; we look for dark, relatively large objects on a plane of limited area. Meteorite searches usually involve a series of transects of a suspected collection area, with a definite spacing between searchers and systematic coverage of an entire region. Empirical models exist for analogous searches, such as those used to estimate whale populations in the ocean, traffic flow on highways, and deposition of garbage around receptacles [26-30]. For this study, a theoretical model has been developed, based on common transect search techniques and models of human visual acuity. Visual acuity is defined as the ability of an observer to visually distinguish objects from background under perfect conditions, such as a bright light on a dark background, or vice-versa. The average person can detect objects with an angular size of about 2 minutes of arc. Because the ability to detect a target is controlled by angular dimensions rather than true size, acuity defines an inverse geometric relationship between the size of an object and the distance at which it is visible; i.e., the farther away something small is, the harder it is to see. Assuming that targets are randomly distributed within a transect of known width, the likelihood that a target of a given size will be found can be calculated. Simulated searches based on these principles, and incorporating conditions specific to ANSMET search procedures (linear, partially-overlapping transects of an entire icefield) suggest that meteorites smaller than approximately 8 g are more likely to be lost than larger specimens (Fig. 3). In addition, the probability of even the smallest specimens is never zero, because it may fall directly in the center of the search transect. For the combined model the calculated probability of particle loss due to searching is applied to the initial supply after wind losses have been incorporated. The combined model / Conclusions: In the combined model the supply functions are treated as a background of meteorite specimens upon which the loss phenomena are immediately superimposed. For most Antarctic icefields this is probably an accurate description of the meteorite collection process. The vast majority of meteorites appear to have resided on stranding surfaces for tens of thousands, if not hundreds of thousands of years, while wind losses occur on a yearly basis and searching occurs over a period of weeks [31]. Figure 4 is a comparison of the combined model and AAM size distributions. The model mimics the observed distribution quite well, within nominal statistical significance. The combined model is the product of a tremendous number of variables with varying influence on the results. Of these variables, it appears that power-law slope, intercept, and the wind speed distribution are most important. In addition, although most simulations run to date provide a single best- fit model, it is not apparent that any single solution is unique. As a result it is not prudent to assume that the model can be used to infer that the values chosen for these variables are valid, or used to calculate an age for the ice surface. However, the ability of the model to produce size distributions similar to those observed does suggest that the various supply and loss phenomena do interact in the manner described above. In addition, the strong dependence of the model on the variables describing power-law supply (presumeably a measure of the age of the icefield) and local wind speed distributions suggest that these factors have a great deal of influence on the size of meteorites that will be found on an icefield. ACKNOWLEDGEMENTS: This work was supported by NSF grant OPP 91-175-58. References: [1] Harvey R.P. and Cassidy W.A. (1989) Meteoritics 24, 9- 14. [2] Gault D. E., Shoemaker E. M. and Moore H.J. (1963) NASA Technical Note D-1767, 39 pp. [3] Hartmann W.K. (1969) Icarus 10, 201- 213 [4] Brown H. (1960) JGR 65, 1679-1683. [5] Dohnanyi J.S. (1972) Icarus 17, 1-48. [6] Hughes, D.W. (1980) In Solid Particles in the Solar System (eds. I. Halliday and B.A. McIntosh), pp. 207-210. IAU, Boston. [7] Greenberg R., and Chapman C.R. (1983) Icarus 55, 455-481. [8] Huss G.R. (1990) Meteoritics 25, 41-56. [9] Zolensky M.E., Wells G.L. and Rendell H. (1989) Meteoritics 25, 11-17. [10] Cassidy W.A. and Harvey R.P. (1991) Geochim. Cosmochim. Acta 55, 99-104. [11] Zolensky M.E., Rendell H. M., Wilson I. and Wells G.L. (1992) Meteoritics 27, 460-462. [12] Ugolini, F.C. (1986) In Rates of Chemical Weathering of Rocks and Minerals (Colman, S. M. and Dethier D. P., eds.) Academic Press. 603 pp. [13] Marvin U.B. (1989) Smiths. Contr. Earth Sci. 28, 113-120. [14] Scott E.R.D. (1984) Proc. 9th Symp. Ant. Mets., Mem. NIPR 35, 102-125. [15] Gooding J.L. (1986) Geochim. Cosmochim. Acta 50 2215-2223. [16] Buchwald V.F. and Clarke R.S., Jr. (1989) Am. Mineralogist 74, 656-667. [17] Harvey R.P. and Score R. (1992) Meteoritics 26, 343-344. [18] Lipschutz M.E. (1982) Smiths. Contr. Earth Sci. 24, 67-69. [19] Schultz L., Weber H.W. and Begemann F. (1990) LPI Tech. Rept 90-01, LPI, Houston, pp 81-82. [20] Velbel M.A., Long D.T. and Gooding J.L. (1991) Geochim. Cosmochim. Acta 55 67-76. [21] Bagnold, R.A. (1960) The Physics of Blown Sand and Desert Dunes. Methuen & Co. Ltd., London. 265 pp. [22] Miller M.C., McCave I. N. and Komar P.D. (1977) Sedimentology 24, 507-527. [23] Greeley R. and Iversen J.D. (1985)Wind as a Geologic Process on Earth Mars, Venus, and Titan. Cambridge University Press. 333 pp. [24] Pye K. (1987) Aeolian Dust and Dust Deposits. Academic Press, London. 334 pp. [25] Bromwich D.H. and Stearns, C.R. (1994) Antarctic Meteorology and Climatology: Studies Based on Automatic Weather Stations. Antarctic Research Series 61, AGU, Washington D.C., 208 pp. [26] Burnham K.P., Anderson D.R. and Laake J.L. (1980)Wildlife Monograph 72, J.Wildlife Man.44. [27] Quinn T.J. II and Gallucci V.F. (1980) Ecology 61, 293-302. [28] Drummer T.D. and McDonald L.L. (1987)Biometrics 43, 13-21. [29] Otto M.C. and Pollock K.H. (1990)Biometrics 46, 239-245. [30] Quang P.X. (1991) Biometrics 47, 269-279.[31] Nishiizumi, K. (1994) (this volume). Wright I. P.* Grady M. M. Pillinger C. T. The Acquisition of Martian Sedimentary Rocks: For the Time Being, Collection as Meteorites from Terrestrial Desert Areas Represents the Best Hope In a recent paper, Wright and Pillinger [1] have discussed some of the problems of identification that would arise if a sedimentary rock of martian origin were to land on Earth as a meteorite (i.e. a sample that had been ejected from Mars and transported to Earth in the same way as its basaltic counterparts, the SNC meteorites). In a sense, this is not a newly recognised problem; the notional difficulties of recognising the origins of meteorites, especially those with planetary affinities, have been outlined by, amongst others, Cross [2] and Nininger [3]. The confusion over assigning origins is further highlighted by the fact that, at one time, some rather illustrious scientists had postulated that meteorites in general may have had a lunar origin [e.g. 4]. Presumably the Moon was imagined to be a more complex place than it turned out to be. But these are thoughts from a previous era, being documented prior to the return of lunar samples by the Apollo and Luna missions of the late 1960s and early 1970s, and before the recent golden age of meteoritics when so many important advances have been made. There is now general agreement that most meteorites have origins from within the asteroid belt; indeed, certain asteroids have been proposed as the parent bodies of particular meteorite groups [e.g. 5,6]. It is thus ironic that meteorites of lunar origin, although comparatively rare, have now been identified [e.g. 7]. In this context the parent body can be considered to be more of a planetary nature rather than an asteroid. But, as planetary bodies go, the Moon is relatively uncomplicated, and so any fragments which arrive on Earth will be relatively easy to identify. In consequence, recognition of lunar meteorites should in the future be fairly straightforward. Indeed, one has been collected from Australia. Note however that it is still possible to misidentify a lunar specimen [e.g. 8], especially when there is pressure to force meteorites into an established classification scheme [e.g. 9]. A far more challenging prospect for those interested in studying planetary meteorites is the quest for samples of sedimentary origin. Throughout the history of meteoritics, materials of this nature have been documented many times; in fact, they are so well established that the different categories have names [2,10], e.g. "amathosites" and "calcarites" are sandstone and limestone meteorites respectively. However, it has to be concluded that whatever samples are currently represented in the world's collections, they are most probably pseudometeorites (i.e. samples of terrestrial origin which, to some people at least, look like meteorites on account of their water- rounded surfaces, or coatings of desert varnish, etc.). The celebrated calcarite, Bleckenstad [11], is purported to contain fossils. But, a terrestrial rock could still be a meteorite. Indeed, Melosh and Tonks [12] have shown that some impact-ejected terrestrial materials could reside in space for 5 million years or so, before returning to the surface - these samples might be observed to fall, have a fusion crust, contain cosmogenically produced nuclides etc. It seems likely, though, that meteorite curators through the ages, having been presented with a fossiliferous sedimentary rock, would probably not have been able to assess the true nature of such a terrestrial meteorite (notwithstanding the protestations of the owner). How many of these types of sample have been returned to sender? More pertinently perhaps, how many such samples are still preserved in traditional collections? In some ways, recognition of terrestrial meteorites is only of academic interest; their main importance would be to document a past impact event. In contrast, meteorites from Mars are extremely valuable. Until appropriate samples can be returned to Earth by space missions, meteorites of martian origin represent the only way of conducting laboratory-based analyses of the planet. Ten samples of the martian crust have been recognised on Earth in the form of SNC meteorites (the evidence for a martian origin can be found in Wood and Ashwal [13] and McSween [14]). But as yet we have no samples of martian sedimentary material. This is certainly not because they do not exist, but is more likely to be because we have not yet learned to recognise them on Earth. What are currently represented on Earth in the form of SNC meteorites are samples of volcanic origin. Most of these materials are relatively fresh examples of shallow intrusive or extrusive basaltic rocks. Indeed, this has been, perhaps subliminally, one of the criteria used for their identification. That the samples are not completely unaltered was first documented by Ashworth and Hutchison [15] and Bunch and Reid [16]. However, the true relevance of the alteration products was not appreciated until the work of Carr et al. [17], who proposed that minerals such as carbonates were martian weathering products produced during their near-surface sojourn. Having established this fact, it became clear that the operation of martian surface processes could be investigated through detailed analysis of these trace weathering products [e.g. 18,19]. However, the overall picture is far from clear; for instance, it has subsequently been discovered that the sample with one of the highest contents of weathering products (EET A79001), also apparently contains high levels of 14C, which seems to suggest a terrestrial origin for some of the carbonates in the sample [20]. Thus far this issue has not been resolved. On the other hand, the most recently identified SNC meteorite (ALH 84001) contains widely distributed, sub-millimetre concentrations of carbonates, which are compositionally zoned, and "clearly formed prior to arrival on Earth" [21]. ALH 84001 was originally classified as a diogenite, a basaltic achondrite of asteroidal origin [21]; in fact, this "new" SNC meteorite is likely to prove pivotal in our understanding of martian surface processes. The carbonates appear to have been formed by hydrothermal alteration at temperatures of about 700 degrees C. This begs the question: if the rock had been more extensively altered, would it have ever been identified as a martian meteorite? On the basis of the surface features on Mars, which can be observed in the photographic records of Mariner 9 and the Viking Orbiters, it is apparent that the geomorphology of Mars has been shaped by fluvial activity, winds, ancient oceans, etc. Models of surface evolution contend that there should be extensive carbonate deposits on Mars, perhaps buried within the regolith. Any impact event which has the power to eject samples of relatively fresh basaltic materials (i.e. from some depth within the crust) must also liberate representatives of the surface rocks. Let us consider the example of a carbonate deposit. If this is successfully transported to the surface of the Earth to arrive as a meteorite, it is interesting to speculate on the appearance of the sample. For instance, it will not have a traditional fusion crust (one of the classical pieces of evidence used to assess the validity of a meteorite). This is because as the sample falls through the Earth's atmosphere, heating at the surface of the sample causes decrepitation of carbonate minerals to a mineral oxide and gaseous carbon dioxide (the gas being lost instantly). In other words, in the case of calcite, CaCO3 would be converted to CaO and CO2. Since CaO has a melting point of 2850 degrees C, the temperatures experienced during infall would never be sufficiently high to enable the production of a glassy fusion crust. Furthermore, upon arrival at the Earth's surface, the CaO would react with atmospheric or meteoric water to produce Ca(OH)2, which in turn would absorb CO2 from the air to form CaCO3. Thus, at no time would the sample look like a classic meteorite. In conclusion, it is probable that recognition of this hypothetical meteorite could only be established by an observation of the fall, and this would have to be by a credible witness (would a member of the public be taken seriously?). If sedimentary martian meteorites do exist they will be present on the Antarctic ice fields. Since it will not be possible to recognize them for what they are in the field, the only way to ensure their collection is to return all samples from within a particular area (as advocated by Huss [22]). The daunting task of positive identification must utilise a rapid screening technique, preferably geochemical analysis of some sort. The most obvious and desirable approach would be to use oxygen isotope determination, since this can be used to identify unambiguously most materials that are not terrestrial in origin (the exceptions being lunar samples, enstatite chondrites, and aubrites). On the other hand, a method specific to limestones, might simply involve carbon isotope analysis. Note that sedimentary meteorites will also be present in the hot deserts, but their retrieval is more problematic - it is not possible to conceive of collecting all rock samples from a particular area. However, members of collection parties can assist in the effort by being aware of materials that seem out of place (sandstones on a limestone platform, for instance). Thinking to the future, it is currently proposed to send mass spectrometers and extraction systems to Mars [23] and a comet [24], for the purpose of making remote stable isotopic measurements. If such instrumentation is developed, its low mass (2-5 kg) may enable the apparatus to be deployed in the field, so that potentially interesting materials could be identified in-situ. In this way collection parties would not need to use industrial-scale earth-moving equipment to transport a field season's samples back to curatorial facilities. In either case, the desired samples of the martian surface would be uncovered several decades before they can be returned to Earth by spacecraft. References: [1] Wright, I.P. and Pillinger, C.T. (1994), Phil. Trans. Roy. Soc. Lond., (in press); [2] Cross, F.C. (1947), Pop. Astron., 55, 96-102; [3] Nininger, H.H. (1967), Meteoritics, 3, 237-251; [4] Urey, H.C. (1959), J. Geophys. Res., 64, 1721-1737; [5] McCord, T.B. et al. (1970), Science., 168, 1445-1447; [6] Cruikshank, D.P. et al. (1991), Icarus, 89, 1-13; [7] Marvin, U.B. (1983), Geophys. Res. Lett., 10, 775-778; [8] Schwarz, C. and Mason, B. (1988) Antarctic Meteorite Newsletter, 11, 21; [9] Lindstrom, M.M. et al. (1994), Lunar Planet. Sci., XXV, 797-798; [10] Graham, A.L. (1985), Catalogue of Meteorites, British Museum (Natural History), 460 pp; [11] Wickman, F.E. and Uddenberg-Andersson, A. (1982), Geol. Foren. Forhan., 104, 57-61; [12] Melosh, H.J. and Tonks, W.B. (1993), Meteoritics, 28, 398; [13] Wood, C.A. and Ashwal, L.D. (1981), Proc. Lunar Planet. Sci., 12B, 1359-1375; [14] McSween, H.Y. (1985), Rev. Geophys., 23, 391-416; [15] Ashworth, J.R. and Hutchison, R. (1975), Nature, 256, 714-715; [16] Bunch, T.E. and Reid, A.M. (1975), Meteoritics, 10, 303-315; [17] Carr, R.H. et al. (1985), Nature, 314, 248-250; [18] Clayton, R.N. and Mayeda, T.K. (1988), Geochim. Cosmochim. Acta, 52, 925-927; [19] Wright, I.P. et al. (1992), Geochim. Cosmochim. Acta, 56, 817-826; [20] Jull, A.J.T. et al. (1992), Lunar Planet. Sci., XXIII, 641-642; [21] Mittlefehldt, D.W. (1994), Meteoritics, 29, (in press); [22] Huss, G.I. (1977), Meteoritics, 12, 141-144; [23] Wright, I.P. et al., (1992), LPI Tech. Rpt., 92-07, 19; [24] Wright, I.P. and Pillinger, C.T. (1993), Ann. Geophys., 11, C472. Thursday, July 21, 1994 COLLECTIONS 8:30 - 9:10 a.m. Chair(s): R. S. Clarke Lindstrom M. M.* Score R. Populations, Pairing, and Rare Meteorites in the U.S. Antarctic Meteorite Collection INTRODUCTION. Meteoriticists have long known that the populations of various meteorite types are different for falls and finds (1). This is attributed to the preferrential collection of iron meteorites as finds because they are dense and easier to distinguish from terrestrial rocks. It was hoped that collection of Antarctic meteorites on ice would be less biased and that Antarctic meteorites would be more representative than other meteorite finds. The last few years have seen a major debate about whether Antarctic meteorites represent different populations than meteorite falls (2,3). We present a review of the populations of meteorites in our collection, pairing of Antarctic meteorites, and the abundances of rare meteorites. We show that our best estimate of Antarctic meteorite populations is very similar to that of falls, except that rare meteorites are more abundant in Antarctic meteorite collections, presumably because they are small and hard to find in other environments. POPULATIONS. The populations of major meteorite types for non-Antarctic falls, finds, and Antarctic finds are given in Table 1. The first three populations are based simply on the total numbers of meteorites. This overall distribution of Antarctic meteorites is more similar to that of non- Antarctic falls than to finds, however the Antarctic population is enriched in ordinary chondrites and depleted in achondrites, stony-irons and irons relative to that of falls. This comparison is flawed by the uncertainty in the number of individual meteorites that the Antarctic meteorites represent. The Antarctic data do not subtract for the possible pairing of two or more meteorite fragments (see below). To avoid this problem, some investigators (6,7) have chosen to compare mass distributions like those in columns 4 and 5. However the simple mass distributions listed here are scewed by the inclusion of large Antarctic irons not found on ice as part of the thorough meteorite search (6,8). Analysis and modelling of the meteorite mass distributions (6,7) concluded that there was no significant difference between falls and Antarctic finds. However, Lipschutz and colleagues (2, 9) have argued based on the numbers of meteorites, that differences in populations are significant. As part of this debate, we offer a simple method of estimating the numbers of Antarctic meteorites that includes pairing corrections for meteorite showers. Table 1. Populations of meteorite types. Percentages of meteorite types are compared for US Antarctic meteorites (4) and non-Antarctic falls and finds (5). Meteorite Non-Ant. Non-Ant. US Ant. Non-Ant. US Ant. US Ant. type falls # finds # finds # falls wt. finds wt. # P=5 Ord. chond. 79.5 49.4 90.2 72.0 77.0 79.5 Carb.chond. 4.2 1.9 3.6 3.1 0.8 5.2 Enst. chond. 1.6 0.7 1.1 nd 0.2 1.7 Achondrite 8.3 1.4 3.4 9.4 2.7 8.5 Stony-iron 1.2 3.6 0.4 4.1 1.6 0.9 Iron 5.1 43.0 1.2 11.2 17.6 4.3 Total met. 830 1588 5702 14904kg 1902kg 1294 Figure 1. Meteorite populations by number. This diagram compares populations for falls, finds and Antarctic meteorites and shows that Antarctic meteorites are more like falls than finds. PAIRING. Meteorites often fall as showers of a few to many fragments. When this happens most places in the world, all fragments are grouped together as the same meteorite. In the Antarctic meteorites from many different falls are concentrated together (8) and it is not simple to determine which ones are from the same fall, or paired specimens. The problem of pairing in Antarctic meteorites has been discussed by several authors (10-12) who have concluded pairing would reduce the number of Antarctic meteorites by factors of 2-10, but most likely 2-6. It is relatively simple to evaluate pairing among the less common meteorite types. Meteorites that are similar to each other and collected in close proximity are most likely from the same fall. Our database of US Antarctic meteorites (4) includes pairing estimates which are known with varying certainty (11). There are 150 pairing groups among the 5700 classified meteorites, with 2-678 meteorites per group. There are, however, only 16 groups with over 10 meteorites and of these only 5 groups have over 25 meteorites. Among the less common meteorites, the number of meteorites per group exceeds 10 in only a few cases: the 20 ALH aubrites, 14 ALH eucrites, 34 ALH CM2, 16 EET CR2, 47 EET CK5 carbonaceous chondrites, and 21 PCA EH3 enstatite chondrites. The average number of specimens per group for the 67 pairing groups of the less common meteorites is 5, in the range of previous estimates of pairing ratios. It is clear that the number of meteorites in a group increases with the the total mass of the group. Among ordinary chondrites pairing is much less certain, but there are some large pairing groups for the ALH L3, QUE L5 and EET L6 chondrites. It is much harder to estimate pairing among H compared to L chondrites, and fewer H pairing groups. We do not believe that this is accurate, but that there are unidentified pairings among H chondrites. Huss (7) proposed the existence of a large ALH H5 pairing group based on mass distributions. In the absence of accurate pairing data for ordinary chondrites, we feel that it is best to use the average paring ratio of 5 from the less common meteorites. When the actual pairing-corrected numbers are used for the five less common meteorite types, and a pairing ratio of 5 is used for ordinary chondrites, the populations of Antarctic meteorites are as shown in the last column of Table 1. This population is remarkably similar to that of non-Antarctic falls as given in the first column. We see no evidence that the overall population of meteorite types has changed over time. Figure 2. Pairing of Antarctic meteorites. This histogram of the number of meteorites per pairing group shows that the average number of Antarctic meteorites per pairing group is 5. Figure 3. Meteorite populations by number, pairing corrected. The populations of Antarctic meteorites, when corrected for pairing, is identical to that of falls. RARE METEORITES. There is, however, strong evidence that the abundance of rare or anomalous meteorites is significantly higher among Antaractic meteorites (3,13-15). Table 2 lists the Antarctic meteorite populations of each of the meteorite subtypes. The last subtype in each major type is that for rare or unusual meteorites. These generally include several different types of meteorites, as for example, the achondrites include acapulcoites, angrites, brachinites, lunar and martian meteorites. Among non-Antarctic meteorites this subtype would include only a few specimens and not amount to more than 1-2%. For achondrites, stony-irons, irons, and carbonaceous chondrites the abundance of rare meteorites is 10-20%! This surprising abundance of rare meteorites in Antarctica is attributed to the fact that it is easier to find small meteorites on ice than on land. Most rare Antarctic meteorites are small specimens (<30g). An exception to this is the planetary meteorites. Although lunar meteorites are found almost exclusively in Antarctica, and most lunar meteorites are <30g, 3 of the 11 are ~500g. Martian meteorites are generally large. Only one of the 4 Antarctic meteorites is small, the others being, 500g, 1900g and 7900g. It would appear that a difference in impact dynamics makes these rare planetary meteorites larger specimens. Table 2. Populations of US Antarctic meteorite subtypes. Percentages are based on pairing-corrected numbers of meteorites, except for ordinary chondrites which are uncorrected. Achondrites % Stony-irons % Irons % aubrites 4.5 mesosiderites 73 grouped 70.4 HED 61.8 pallasites 9 group-anom 8.5 ureilites 18.2 lodranites 18 ungrouped 21.1 primitive/planet 15.5 Carbonaceous % Enstatite Ch. % Ordinary Ch. % C2/CM 53.7 EL 36.4 H 49.3 CO3/CV3 16.4 EH 31.8 L 46.6 CK4-6 14.9 E unclassified 27.3 LL 4.0 C anomalous 10.9 E anomalous 4.5 OC anomolous 0.1 REFERENCES. (1) Mason B. (1962) Meteorites. John Wiley and Sons. (2) Dennison J., Lingner D., and Lipschutz M. (1986) Nature 319:390-393. (3) Koeberl C. and Cassidy W. (1991) Geochim. Cosmochim. Acta 55:3-18. (4) Grossman J. (1994) Meteoritics 29, 100-143 and Ant.Met.News. (1994), 17(1). (5) Graham A.L., Bevan A.W.R., and Hutchison R. (1985) Catalogue of Meteorites. 4th ed. British Museum (Natural History). (6) Cassidy W.A. and Harvey R.P. (1991) Geochim. Cosmochim. Acta 55, 99- 104. (7) Huss G. (1991) Geochim. Cosmochim. Acta 55:105-111. (8) Cassidy W., Harvey, R., Schutt J., Delisle G., and Yanai K. (1992) Meteoritics 27:490-525. (9) Lipschutz M. (1989) Workshop on Differnences between Antarctic and non-Antarctic Meteorites, LPI, 59-61. (10) Graham A.L. and Annexstad J.O.(1989) Antarctic Sci. 1: 3-14. (11) Scott E.R.D. (1989) Smithsonian Contrib. Earth Sci. 28: 103-111. (12) Ikeda Y. and Kimura K. (1992) Meteoritics 27:435-441. (13) Marvin U.B. (1983) New Scientist 17, 710-715. (14) Clarke R.S. (1986) International Workshop on Antarctic Meteorites, LPI, 28-29. (15) Wasson J. (1990) Science 249:900-902. Geiger T.* Bischoff A. Meteorite Find Locations, Shock Classification, and Pairing of 464 Meteorites from the Sahara and the Mineralogical and Chemical Characterization of Rare Types 470 meteorites were collected between 1989 and 1993 mainly from the Algerian part of the Sahara. 464 samples have been studied at the Institute of Planetology in Munster; some samples were sold by the finder before classification and were not available for studies. The most important find locations of these samples are given in Fig. 1a. In addition, the location of the Daraj area is included, where 54 meteorites were recovered in the past. Most meteorites (319) were found in the Acfer region about 30x100 km in size (Fig. 1b). While the masses of most meteorites from Antarctica, Roosevelt County, and the Nullarbor Plain are on the order of 10-100 g, meteorites found in the Sahara are generally larger. Most meteorites are weighing between 100 and 1000 g (Fig. 2). If pairing is not considered, 433 ordinary, 21 carbonaceous, and 3 enstatite chondrites occur among the 464 meteorites. One meteorite belongs to the new group of Rumuruti (R) chondrites [1,2]. In addition, one ureilite, three mesosiderites, and two iron meteorites were identified. We have tried to solve the pairing problem by considering the find location, the degree of weathering, the degree of shock metamorphism, and the mineral chemistry (olivine, pyroxene). In Tab. 1 the number of meteorites of different meteorite classes and their frequency distributions are listed along with the frequencies of finds from Antarctica, the Nullarbor desert, and of meteorite falls. The distribution of the 297 ordinary chondrites in various classes (H,L,LL) and petrologic types (3-6) is given in Tab. 2. The degree of shock metamorphism of the stony meteorites was obtained according to the classification system of Stöffler et al. [3]. The results for the ordinary chondrites are given in Fig. 3. From the 21 carbonaceous chondrites 12 samples belong to the CR chondrite Acfer 059/El Djouf 001 [4] and three samples to the CH chondrite Acfer 182 [5]. All others are unpaired. Thus, eight carbonaceous chondrites are among the set of meteorites from the Sahara. Rare meteorite types: Acfer 182 is chemically, texturally, and mineralogically similar to ALH85085 [5] and PCA 91467 [6]. Considering their affinity to carbonaceous chondrites, their high bulk iron content and the high metal abundance they were designated as CH chondrites [5]. In Ca,Al-rich inclusions from Acfer 182 grossite is a very abundant phase [7,8]. Acfer 217 has chemical and mineralogical properties very similar to Rumuruti and Carlisle Lakes [1]. It is a regolith breccia with abundant olivine (~72 vol%), which has a high Fa-content of 37-39 mol%. With the meteorite Rumuruti, the first fall of this type of chondrite (eight members) is known; therefore, this group has been designated as the R chondrites [2]. Acfer 094 is a uniquely primitive carbonaceous chondrite, which has more diamond and SiC than any other specimen studied [9]. Based on the mineralogy, chemistry, and oxygen isotope characteristics it is not possible to distinguish the meteorite unambiguously between CO3 and CM2 chondrites. It is suggested that Acfer 094 may be the first CM3 [10]. The other carbonaceous chondrites from the Sahara were briefly characterized [11]. Also, the ureilite Acfer 277, the H3-6 chondrite regolith breccia Acfer 111 [12], and the EL- chondritic melt rock Ilafegh 009 [13] are of great interest for meteorite research. References: [1] Bischoff A. et al. (1994) Meteoritics 29, 264-274; [2] Schulze H. et al. (1994) Meteoritics 29, 275-286; [3] Stoffler D. et al. (1991) Geochim. Cosmochim. Acta 55, 3845-3867; [4] Bischoff A. et al. (1993) Geochim. Cosmochim. Acta 57, 1587-1604; [5] Bischoff A. et al. (1993) Geochim. Cosmochim. Acta 57, 2631-2648; [6] Bischoff A. et al. (1994) Meteoritics 29, 444; [7] Weber D. and Bischoff A. (1994) Eur. J. Mineral. 6, 591-594; [8] Weber D. and Bischoff A. (1994) Geochim. Cosmochim. Acta 58 (in press); [9] Newton J. et al. (1994) Meteoritics (submitted); [10] Bischoff A. and Geiger T. (1994) LPS XXV, 115-116; [11] Geiger T. and Bischoff A. (1992) Meteoritics 27, 223; [12] Pedroni A. and Weber H.W. (1991) Meteoritics 26, 383-384; [13] Bischoff A. et al. (1992) LPS XXIII, 105-106; [14] Koeberl C. et al. (1992) Geowissenschaften 8, 220-225; [15] Bevan A.W.R. (1992) Records of the Australian Museum; [16] Huss G.R. (1991) Geochim. Cosmochim. Acta 55, 105-111. Fig. 1: a) Main find locations in the Sahara; EA = El Atchane; Tanezr. = Tanezrouft; HAH = Hammadah Al Hamra; b) meteorite finds in the Acfer- Aguemour region. Fig. 2: Number of meteorites vs. mass of meteorite falls and finds from different areas (after [14,16]). Fig. 3: Shock classification of ordinary chondrites from the Sahara; data are corrected for pairing. Table 1: Number and frequency (%) of meteorites from the Sahara (corrected for pairing) and other locations [15]. Table 2: Classification of ordinary chondrites from the Sahara (corrected for pairing). Thursday, July 21, 1994 WEATHERING EFFECTS 9:10 - 11:20 a.m. Chair(s): M. E. Zolensky Ash R. D.* Pillinger C. T. The Fate of Meteoritic Carbon in Hot and Cold Deserts INTRODUCTION. As a result of the large influx of meteorites into laboratories from hot and cold deserts many previously rare meteorites are becoming more available for study and new and unique meteorites have been recovered. Although the samples recovered often show remarkably little weathering, considering their terrestrial ages, they are not pristine and both mineralogical changes and oxidation are ubiquitous. The chemical changes show variation between localities, for example halogens appear depleted in the Sahara [1] whereas they are commonly enriched in the Antarctic [2]. Many other elements show similar variations with enrichments in some deserts, depletions in others. In this paper we shall review the effects of long terrestrial residence times upon carbon contents and occurrence of meteorites from three desert localities; the cold deserts of Antarctica, the semi-arid region of Roosevelt County, USA and the hot desert of the Sahara, North Africa. The form of preservation and the climate play an important part in the style of weathering and the potential contaminant species. In the Antarctic the meteorites have been preserved in the ice but also spend considerable time on meteorite stranding surfaces, in the Sahara meteorites reside on the surface of limestone plateaux and in Roosevelt County by burial in sand and soil and subsequent uncovering by wind ablation. The two most common forms of carbon in meteorites are organics and carbonates, both of which are also present in high abundances on the surface of the Earth which leads to potential problems with contamination. Other, minor carbonaceous components, such as presolar materials (diamond, graphite, SiC), appear unaffected by residence in deserts. FALLS and FINDS. The first studies of the effects of terrestrial residence on carbon were carried out on whole rock samples using bulk techniques. Early work on the carbon abundances of ordinary chondrites showed that there was no difference in the median values for falls and finds of the H and L group chondrites [3]. However if the mean values of the L and LL groups from this study are considered there is a two fold depletion in the carbon contents of the finds over falls, something observed in all petrologic grades. The falls in this study were not selected from any particular environment but the majority of the finds were from the farmlands of the midwest USA so had generally been subject to burial in prairie land soil. A later study of several specimens of the meteorite Holbrook which had been collected at various times after its fall (in 1912) and a three fold increase in carbon content was found between samples collected immediately after fall (517ppm) and one collected 19 years later (1560ppm). However the weathering effects appear to be variable as a sample collected a further 37 years later had an intermediate carbon content (1125ppm) [4]. Neither of the above studies determined the isotopic compositions of the carbon, nor the nature of the material lost or gained. Later studies of ordinary chondrites have included data on carbon contents and isotopic compositions (Fig. 1). This has made the identification of weathered materials and introduced species easier. It has been found that the effects of weathering on ordinary chondrites tends to increase the carbon content, but for different reasons. ORGANICS. The carbon budget of primitive chondrites is dominated by a macromolecular highly insoluble organic material. This constitutes up to 90% of carbon in carbonaceous chondrites, contributing less to metamorphosed carbonaceous and ordinary chondrites where graphitic material becomes dominant. The organic material found in carbonaceous chondrites is severely affected by the conditions found in the hot deserts of the Sahara. The temperature of the meteorites on the desert floor (>80 degrees C [5]) is high enough for the destruction of the macromolecular material, probably by a process akin to catalytic hydrolysis. These reactions are enhanced by the presence of alkali groundwaters and involves the cleaving of carbon-carbon bonds in the macromolecular organic material to produce soluble and/or volatile organics which are then lost to the environment. The Saharan carbonaceous chondrites show up to 80% loss of their organic carbon content compared with other, non-Saharan members of the group. Carbonaceous chondrites which have undergone more parent body heating, such as the C3s, show less carbon depletion than the lower temperature chondrites due to the greater resistance to degradation through weathering due to their more graphitic nature [6]. Unlike the surface preservation of meteorites in the Sahara, meteorites from the semi-arid region of Roosevelt County (New Mexico and West Texas) have spent the majority of their terrestrial residence buried in the soil. The result of this is that they have gained organic carbon [7] and the longer the terrestrial residence the more weathered the samples become [8] and the more carbon-rich the meteorites become (FIG. 2 - the meteorite of weathering grade C with the high carbon content is RC 075[see below]. For simplicity the high indigenous carbon content has been subtracted from the total leaving just the carbon attributable to weathering). This carbon can be separated analytically from the indigenous, meteoritic carbon by stepped combustion, with the terrestrial material combusting before the meteoritic carbonaceous compounds. This was carried out upon the primitive H chondrite RC 075 which has highest carbon content of any ordinary chondrite of ca. 1.5%. Using stepped combustion it was found that more than half of the carbon in this sample has a terrestrial origin, with a delta^13C of ca.-25 per mil [9]. This is the typical isotopic composition for terrestrial plant matter (range -24 per mil to -34 per mil [10]) and for recent sediments (range -20 per mil to -27 per mil [11]). There is no evidence for any destruction of the meteoritic carbonaceous material, although the absence of any carbonaceous chondrites from Roosevelt County makes this difficult to assess, but the organic material from RC 075 appears isotopically unchanged. Analyses of Antarctic chondrites indicate that the organic materials are unaffected by their residence in the polar ice. Stepped combustion shows that there is some increase in the lowest temperature material (see below and NEWTON et al., this volume) but that the organic peak in the combustion remains isotopically and quantitatively indistinguishable from observed falls. CARBONATES. Carbonaceous chondrites contain up to 5400ppm indigenous carbonate. It is formed by parent body hydrothermal processes and is generally isotopically distinct from terrestrial material, being isotopically heavy with delta^13C values up to +81 per mil [12]. Although seemingly absent from ordinary and enstatite chondrites carbonates are also present in the SNC meteorites as a Martian weathering product or hydrothermal deposit. Carbonaceous chondrites from the Sahara show evidence for the precipitation of up to 10000ppm carbonate. This has been shown to be entirely calcitic in composition [13], whereas all the magnesian carbonate in the meteorites has an isotopic composition commensurate with an extraterrestrial origin. The calcium carbonate is introduced by groundwater which is carbonate-rich from percolation through limestones before being drawn into the meteorite by osmosis. The evaporation of the water leads to the deposition of carbonates in veins commonly associated with other weathering features such as iron oxide deposits. This is potentially a powerful tool for the destruction of meteorites in the desert environment. In the cold deserts of Antarctica the deposition of evaporitic carbonate takes place to a lesser extent and is of different composition. The majority of the cold desert deposits are hydrated carbonates, such as nesquehonite and hydromagnesite [14]. These have a carbon isotopic composition of ca.+1 per mil and are believed to have been formed from a mixture of atmospheric carbon dioxide and CO2 dissolved in ice water [15]. Carbonates of terrestrial origin are also found in the SNC meteorites [16]. Despite the preservation within sands, which are believed to have been at some stage cemented with a calcareous cement (ZOLENSKY, pers. comm.), the meteorites of Roosevelt County appear to have escaped carbonate contamination. Stepped heating of RC 075, the meteorite from this region most affected by organic contamination, shows that the amount of carbon yielded over the carbonate decomposition range (600- 700 degrees C) is equivalent to less than 500ppm calcite. Visual inspection of some samples from the Nullarbor Plain, Western Australia show evidence for the presence of carbonates from this semi-arid limestone environment. LOW TEMPERATURE CARBON. The presence of high carbon yields at low temperatures (<250 degrees C) in the stepped combustion analyses were first noted in Antarctic ordinary chondrites [17]. The composition of this material was found to be isotopically heavy with respect to the bulk meteorite, with delta ^13C values up to 0 per mil. Analysis of other chondritic meteorites from the Antarctic have shown similar effects, as have meteorites from the Sahara and Roosevelt County. A more detailed study has been undertaken which has shown possible geographical dependence of the degree of low temperature contamination in the Antarctic and it has been suggested that it may be connected with aerosol sea spray (NEWTON et al., this volume). CONCLUSIONS. The carbon content and composition of meteorite finds from both hot and cold deserts is affected by the environment in which they are found (summary see Table 1). Thus care must be taken when interpreting carbon data, particularly from bulk, whole rock samples from these regions. Some effects, such as the organic contamination encountered in Roosevelt County meteorites, can be overcome by the application of suitable analytical techniques, e.g. stepped combustion. Other problems, such as the loss of organic material from hot desert samples and the growth of carbonates in Saharan and Antarctic samples, at present appear more intractable. References: [1] Bischoff et al. (1993) G.C.A. 57, 1587. [2] Langenauer & Krahenbuhl (1993) Meteoritics 28, 384. [3] Moore & Lewis (1967) J. Geophys. Res. 72, 6289. [4] Gibson & Bogard (1978) Meteoritics 13, 277. [5] Hume (1925) Geology of Egypt. Vol. 1. Ministry of Finance, Government Press, Cairo. [6] Ash & Pillinger (1994) Meteoritics in press. [7] Ash & Pillinger (1993) L.P.S.C. XXIV, 43. [8] Jull et al., (1991) L.P.S.C. XXII, 667. [9] McCoy et al. (1993) Meteoritics 28, 681. [10] Smith & Epstein (1971) Plant Physiol., 47, 380. [11] Eckelman et al., (1962) Bull. Am. Ass. Petrol. Geol. 46, 699. [12] Grady et al. (1988) Geochim. Cosmochim. Acta 52, 2855. [13] Grady and Pillinger (1993) E.P.S.L. 116, 165. [14] Jull et al. (1988) Science 242, 417. [15] Karlsson et al. (1991) L.P.S.C. XXII, 689. [16] Jull et al. (1992) L.P.S.C. XXIII, 641. [17] Swart et al., (1983) Meteoritics 18, 137. [18] Grady et al. (1982) J. Geophys. Res.87, A289. [19] Grady et al. (1989) Meteoritics 24, 147. Figure 1, showing a comparison of the carbon content and isotopic composition of ordinary chondrite falls and ordinary chondrite finds, appears here in the hard copy. Figure 2, showing the carbon content of Roosevelt County ordinary chondrites of known weathering grade, appears here in the hard copy. Table 1, showing carbonaceous components and characteristics of desert meteorites, appears here in the hard copy. Bland P. Berry F. J. Pillinger C. T.* Iron-57 Mossbauer Spectroscopy Studies of Weathering in Ordinary Chondrites from Roosevelt County, New Mexico Introduction Meteorite weathering can be regarded as the alteration of original component phases of the meteorite to phases that are more stable at the Earth's surface. On entering the Earth's atmosphere, interaction with the terrestrial environment begins. Meteorites are uniquely placed among geological materials in that they show relatively minor intra-group variations and their terrestrial age can be established by measuring the decay of cosmogenic radionuclides by accelerator mass spectrometry [1]. They are therefore a potential "chronometer" of environmental conditions during their terrestrial residency. Being the most common meteorite type and having a tightly constrained mineralogy [2], ordinary chondrites are ideal candidates for investigating terrestrial weathering products in meteorites. Their relatively high iron content makes them suitable for examination by 57Fe Mossbauer spectroscopy. Arid climate, uniform topography and lack of a concentration/movement mechanism makes it likely that meteorites throughout Roosevelt County were weathered by a common mechanism. Jull et al (1991) showed a correlation in meteorites from Roosevelt County between terrestrial 14C ages and a qualitative weathering scale. We report here on a our examination of a suite of meteorites by 57Fe Mossbauer spectroscopy recovered from Roosevelt County, for which terrestrial ages have been determined [3]. Fresh fall meteorites, Allegan (H5) and Barwell (L6), were used as unweathered standards for the purposes of this study. Meteorite Specimens Meteorites were obtained from the Natural History Museum, London and the Max-Planck-Institut für Chemie, Mainz, Germany. Approximately 0.5-1.0 gram of sample was used, prepared by grinding under acetone to prevent oxidation during crushing, until a homogenised powder was produced. Mössbauer spectra were recorded at 298 degrees K and 77 degrees K with a microprocessor controlled Mossbauer spectrometer using a 57Co/Rh source. Drive velocity was calibrated with the same source and a metallic iron foil. The Mossbauer spectra were fitted with a constrained non-linear least squares fitting program of Lorentzian functions. The fitted lines were integrated to give the relative area intensities of the iron-containing phases in the sample. Results and discussion Components of Mossbauer spectra The Mossbauer spectra recorded from Barwell and Allegan validate the assumption that these samples represent unweathered meteorites: no Fe^(3+) was detected in either of these samples. However, spectral components associated with Fe-Ni metal, troilite, olivine and pyroxene [4] were recognised. The Mossbauer spectra recorded from Roosevelt County meteorites showed similar features but with different intensities of the separate components. In addition, absorptions associated with Fe^(3+)-containing terrestrial corrosion products were observed, with a concomitant reduction in the spectral area of pre-terrestrial meteoritic minerals. Although the Mossbauer parameters of Fe-oxides and oxyhydroxides are similar, they are sufficiently different, particularly when spectra recorded at 298 degrees K and 77 degrees K are compared, to allow the assignment of individual phases. In this study, spectra recorded at room temperature and at liquid nitrogen temperatures indicate the presence of goethite, magnetite, maghemite and akaganeite. Ferrihydrite and lepidocrocite are also tentatively identified. This suite of oxidation products corresponds to that observed by [5] and [6] in an electron microprobe, SEM and XRD study of weathered meteorites. This would suggest a broadly similar corrosion mechanism to that proposed by [5], with akaganeite precipitating at the reaction front as metal goes into solution, and with time and distance from the reaction surface, gradually transforming to maghemite and goethite. Magnetite has been observed as a significant corrosion product in hot desert meteorites by [6], and it is also a common component in the Mossbauer spectra from the Roosevelt County meteorites described in this study. In a similar study of a suite of Antarctic H-chondrites, we found no absorption associated with magnetite in the Mossbauer spectra from these samples. A possible explanation for this may be that in the relatively warmer and more humid environment of a hot desert, the rapid dissolution and oxidation of Fe-Ni (containing Fe^0) may produce a solution around Fe-Ni grains that is saturated with Fe^(2+), before these ions are also oxidised to Fe^(3+). Although the relatively minor proportion of Fe-Ni remaining in Roosevelt County meteorites does not allow such a saturation effect to be observed, it is possible that the presence of these two iron species allows the stability of magnetite, which may then persist in this environment as a metastable reaction product [7]. The oxidation of olivine and pyroxene may give rise to iron oxyhydroxides such as ferrihydrite, lepidocrocite and goethite [8]. Relationship between weathering and terrestrial age The contribution to the Mossbauer spectra from the ferromagnesian silicates olivine and pyroxene has been calculated from computer fitted spectra recorded from Barwell and Allegan, and a suite of Roosevelt County meteorites. The data are plotted against terrestrial age in Figure 1. A clear trend is observed showing a decreasing contribution from these phases with increasing terrestrial age. The linear fit to this data is significant at better than the 99% level, indicating a rate of dissolution and oxidation for ferromagnesian silicates of approximately 5% per 5000 years. This would imply the complete oxidation of these silicates to secondary iron oxides and oxyhydroxides in Roosevelt County ordinary chondrites after around 60,000 years, thus giving an approximate upper limit to the residence time of meteorites such as RCO 48 and RCO 56 for which 14C dating can only provide a minimum terrestrial age [3]. Figure 1, showing the spectral area of ferromagnesian silicates against terrestrial residence time in Roosevelt County ordinary chondrites, Barwell and Allegan, appears here in the hard copy. Interestingly, two of the meteorites which deviate most from this line (RCO 45 and RCO 65) fell at a time when climatic conditions in Roosevelt County, and much of the rest of North America, were changing dramatically [9], [10]: at the end of the last glaciation. A plot of Fe^(2+)/Fe^(3+) (derived by combining the spectral areas of Fe^(2+)-containing phases (olivine, pyroxene and troilite) and dividing by the combined spectral areas of Fe^(3+)-containing phases) against terrestrial age also indicates the possibility of a climatic control on meteorite weathering. The period of the Quaternary over which these meteorites fell has been divided by [11] into three stages, using an oxygen isotope stratigraphy; these stages are shown on Figure 2. It is clear from this diagram that there is some correlation between meteorite weathering over time and the independently measured oxygen isotope stratigraphy, that is itself a function of climatic changes during the late Quaternary. Figure 2 shows a regular increase of Fe^(2+)/Fe^(3+) between 32,300 and 25,900 BP (isotope stage 3), followed by a period of relatively little oxidation, then a drop to lower Fe^(2+)/Fe^(3+) values around 14,000 BP (isotope stage 2) before a steady increase again (isotope stage 1). The meteorite falls in Roosevelt County span a period up to, or exceeding, 50,000 years [3]. As such, the oldest meteorites from this region fell to Earth during the middle of the last glaciation. The period following on from this time (up to approx. 24,000 BP), characterised by [11] as isotope stage 3 (this corresponds approximately to the Middle Wisconsin stage in the U.S. and Canada) involved fairly stable climatic conditions. This period is widely recognised as an episode of general climatic amelioration in North America, involving retreat and decay of the Laurentide and Cordillerian ice sheets [12]. In fact, it has been suggested by [13] and [14] that these ice sheets may have melted completely during the Middle Wisconsin. Although there is little local evidence of what the climate was like in Roosevelt County at this time, the overall picture of climatic amelioration in North America may have produced an environment not dissimilar to that which we observe today, leading to a similar rate of weathering in meteorites that fell during this time to those that fell during the last few thousand years The culmination of the last glacial maximum occurred at 18,000 BP [15]. This period saw the lowest temperatures at any time during the 110,000 years of the last glaciation [16]. It is possible that this depressed the rate of chemical weathering reactions, causing relatively little oxidation in the meteorites that fell around this time. Between the time when RCO 45 and RCO 65 fell, temperatures increased rapidly worldwide [16], leading to melting of glaciers, increased runoff and increased precipitation. Meteorites falling after this period, during which climate was relatively stable, would show more gradual oxidation. Figure 2, showing Ratio of Fe^(2+)/Fe^(3+) against terrestrial residence time in ordinary chondrites from Roosevelt County, appears here in the hard copy. Postulating a climatic control for this difference in weathering rate requires that meteorites with older terrestrial ages must have been protected in some way during the period of increased chemical weathering i.e. samples must 'remember' a previous milder weathering regime, rather than lose the information during periods of increased weathering. [17] found that oxygenated dissolution of iron-rich minerals (in this case olivine) leads to the formation of two surface layers, the inner layer being a Fe^(3+)-Mg silicate which could be protective toward further dissolution. More recently, [18] have observed laihunite (an Fe^(2+), Fe^(3+), Mg silicate) as an alteration product of olivine, forming intergrowths with the original crystal which increase its resistance to weathering. The role of surface species in controlling the low temperature dissolution of minerals has also been emphasised by [19]. If protective stable oxide layers do have a passivating role in the continued oxidation of primary meteorite phases, meteorites may survive periods of more intense weathering, in effect preserving an assemblage of primary and secondary minerals that was produced during a different climatic regime. Such oxides may act to protect meteorites with longer terrestrial ages during times of increased weathering, whilst meteorites falling during such periods, i.e. starting their weathering history, may experience higher rates of oxidation. Conclusions This study has shown that 57Fe Mossbauer spectroscopy is a promising means by which the course of oxidation and the nature of the weathering processes in meteorites may be identified. The weathering of ordinary chondrites in Roosevelt County is seen to be a function of time and of climatic changes that have occurred during the meteorites terrestrial residency. We have also conducted a similar study of Antarctic H-chondrites and these meteorites do not show this effect. It seems likely that, as suggested by [20], time exposed on the ice surface rather than absolute terrestrial age may be the controlling factor in weathering of Antarctic meteorites. References: [1] Jull A. J. T., Donahue D. J. and Linick T. W. (1989) Geochim. Cosmochim. Acta. 53, 2095-2100. [2] Mason B. (1965) Am. Mus. Novitates 223, 1-38. [3] Jull A. J. T., Wlotzka F. and Donahue D. J. (1991) Lunar Planet. Sci. XXII. 667-668. [4] Ortalli I. and Pedrazzi G. (1990) Hyperfine Interactions 57, 2275-2278. [5] Buchwald V. F. and Clarke Jr. R. S. (1989) Am. Miner. 74, 656-667. [6] Buchwald V. F. (1989) LPI Technical Report 90-01, 24-26. [7] Buchwald V. F. (1994) pers. comm. [8] Burns R. G. (1993) GCA 57, 4555-4574. [9] Dyke A. S. and Prest V. K. (1987) Geog. Physique Quaternaire 41, 237-264. [10] Dawson A. G. (1992) Ice age earth: late quaternary geology and climate. London, Routledge. [11] Martinson D. G., Pisias N. G., Hays J. D., Imbrie J., Moore, Jr. T. C. and Shackleton N. J. (1987) Quat. Res. 27, 1-29. [12] Sancetta C., Imbrie J. and Kipp N. C. (1973) Quat. Res. 3, 110-116. [13] Clague J. J. (1981) Geol. Survey of Canada, Paper 80-35, 41pp. [14] Fulton R. J., Fenton M. M. and Rutter N. W. (1986) In V. Sibrava, D. Q. Bowen and G. M. Richmond (eds) Quaternary Glaciations in the Northern Hemisphere, 229-242. [15] Ruddiman W. F. and McIntyre A. (1981) Palaeogeog., Palaeoclim. Plaeoecol. 35, 145-214. [16] Jouzel J., Lorius C., Petit J. R., Genthon C., Barkov N. I., Kotlyakov V. M. and Petrov V. M. (1987) Nature 329, 403-407. [17] Schott J. and Berner R. A. (1983) Geochim. Cosmochim. Acta. 47, 2233-2240. [18] Banfield J. F., Veblen D. R. and Jones B. F. (1990) Contrib. Min. Pet. 106, 110-123. [19] Blum A. and Lasaga A. (1988) Nature 331, 431-433. [20] Gooding J. L. (1986) LPI Technical Report 86-01, 48-54. Crozaz G.* Pyroxene, The Indicator of Pervasive Trace-Element Mobilization in Antarctic Meteorites The discovery of large numbers of meteorites in Antarctica has stimulated many cosmochemical studies; new meteorite types were recognized and groups of rare meteorites were greatly expanded. In addition, the availability of meteorites that fell much earlier (tens to hundreds of thousand years ago) than those represented in the museum collections provided the opportunity to compare these two populations of objects. Of critical importance to these comparisons has been the question of whether compositional differences between meteorites of a given group were generated in their parent body (ies) or whether they reflect the extremely different terrestrial histories experienced by Antarctic and non-Antarctic meteorites. It was first assumed that Antarctic meteorites had remained essentially unchanged in their cold and dry environment but then realized that weathering, though at a slow rate, had been at work. The presence and abundance of rust became and still are the basis for the weathering classification of these objects; they guided the choice of meteorites to be studied by cosmochemists. However, the production of rust depends not only on the length and intensity of weathering but, more importantly, on the amount of Fe-Ni metal and troilite present in the sample. Some meteorites (e.g., eucrites, SNCs...) are metal- and sulfide -poor and thus, even after severe weathering, may not be recognized as altered. In eucrites, there is considerable evidence that the REEs, so widely used in petrogenetic modelling, were disturbed (e.g., [1]). Mittlefehldt and Lindstrom [1] have shown that many of the Antarctic eucrites they analyzed (~60%) have REE patterns with positive (and sometimes negative) Ce anomalies, positive Eu anomalies, and low abundances of the remaining REEs. Their data strongly imply that while in or, most likely, on the ice, eucrites are altered and lose part of their REEs. These authors attributed the Ce anomalies to the partial oxidation Ce^(3+) to Ce^(4+) and the partitioning of the more insoluble Ce^(4+) from the other REEs when dissolution of the major REE carriers in eucrites, the calcium phosphates, occurred. Dissolution of these minerals was promoted by the production of an acid solution while the meteorite was residing on the ice surface. Because the phosphates are located in interstitial material, it was assumed [1] that the major silicate phases were left intact and thus are suitable for petrogenetic modelling. Ion microprobe measurements of REE concentrations in individual grains, made in our laboratory and described below, confirm the importance of phosphate dissolution in altering the whole rock patterns but also demonstrate that other types of meteorites as well as minerals other than phosphates, most prominently and commonly pyroxene, were significantly affected by weathering processes. That the weathering of phosphate tends to leach the REEs leaving Ce preferentially behind was accidentally supported by observations in an acid etched section of the heavily shocked shergottite ALHA77005. The REE patterns of individual merrillite grains (Fig. 1) are identical except for positive Ce anomalies whose sizes correlate positively with the apparent REEs concentrations. A range of REE concentrations is observed (factor of 25) that clearly exceeds the variation expected during closed system crystallization of a melt. In fact, the range of concentrations is not real but an artifact of the Ca normalization used to derive the data. The results imply that Ca and the REEs were all leached but that Ca and Ce were, respectively, the easiest and the most difficult elements to mobilize. In contrast, analysis of an unetched section of the same meteorite did not reveal any Ce anomalous merrillite [2]. In eucrites, there are three factors that facilitate REE leaching from Ca phosphates: grains of these minerals are typically smaller than in other types of objects, they are unusually rich in actinides (which caused extensive radiation damage) and, in meteorites that have undergone a significant shock event, they have acquired numerous shock- induced defects. Leaching of REEs from phosphates not only produces a net loss of these elements but also leads to their redistribution within the meteorite itself. Other minerals, particularly pyroxene, are also affected. In a detailed study of a eucrite [3], we showed that plagioclase, pyroxene and silica can all have Ce anomalies (positive and/or negative). Pyroxene was most affected (38 out of 52 grains had Ce anomalies) because it has an extensive, shock-induced network of microcracks. Conversely, Ce anomalies are rare in plagioclase (in only 2 out of 17 grains) because it lacks a network of defects along which REEs can be mobilized. No single explanation can account for the observations in pyroxene. Whereas the negative Ce anomalies (present in 12% of the analyses) could be due to the addition of REEs (minus Ce) generated by the phosphate dissolution, the positive Ce anomalies (in ~60% of the analyses) suggest that the REE redistribution mechanism is actually much more complex and involves leaching from the pyroxene itself. In this context, it should be noted that the distribution of Ce anomalies in pyroxene is quite heterogeneous, even on a scale of just a few hundred microns. A series of 8 analyses along a single pyroxene grain resulted in REE patterns that have positive, negative or no Ce anomalies, with no obvious trend according to location. Since the analysis of this eucrite (whose whole rock REE pattern has a Ce anomaly), more anomalies have been found in pyroxene from meteorites in which the weathering effects are less obvious (i.e., do not result in a bulk REE pattern with significant Ce depletion or enrichment). These observations are a by-product of our ongoing petrogenetic studies that rely on the use of the ion microprobe. Without any exception, all the meteorites that contain pyroxene with Ce anomalies were found in Antarctica. To date, they include all the Antarctic shergottites, ALHA77005, EETA79001 and LEW88516, which are heavily shocked, and the EH3 chondrite Y-691. Ce anomalies in Y-691 enstatite (Fig. 2) are rare (only 2 examples out of tens of grains) and are presumably due to the oxidation and dissolution of oldhamite (CaS), the major REE carrier in enstatite meteorites. We also found (Fig. 3) a pigeonite in a ureilite from Antarctica with a very pronounced positive Ce anomaly. Thus, it is clear that REE mobilization was not limited to eucrites but affected to various degrees many types of meteorites. In conclusion, weathering in Antarctica not only results in REE loss but also extensive redistribution within the meteorites. Pyroxene, particularly when shocked, is a sensitive indicator of these processes. Therefore, it is risky to assume that the major silicate phases of Antarctic meteorites remained pristine and that they can always be used in petrogenetic modelling. References: [1] Mittlefehldt D. W. and Lindstrom M.(1991) Geochim. Cosmochim. Acta 55, 77-87. [2] Lundberg L.L., Crozaz G. and McSween H. Y. (1990) Geochim. Cosmochim. Acta 54, 2537-2547. [3] Floss C. and Crozaz G. (1991) Earth Planet. Sci. Lett. 107, 13-24. Figure 1, showing apparent REE abundances in five merrillite grains from ALHA 77005, appears here in the hard copy. Figure 2, showing REE patterns with Ce anomalies in enstatite grains from the EH3 chondrite Y-691, appears here in the hard copy. Figure 3, showing REE pattern with a striking positive Ce anomaly in pigeonite from the ureilite ALHA 78019, appears here in the hard copy. Krahenbuhl U.* Langenauer M. Comparison of the Distribution of Halogens in Chondrites from Antarctica and from Western Australia All finds of meteorites not sampled immediateiy after its fall are subjected to alterations either to enrichments due to contamination or to depletions e.g. by leaching [1]. It is important to know how deeply into the interior such processes are affecting the original composition of the investigated specimen. After the investigation of the distribution of the 4 halogens in a large number of H 5 and H 6 chondrites from different locations within Antarctica the goal of this investigation was to extend a similar study to samples from a hot desert. In this study three H 5 chondrites from the Nullarbor Plains, Western Australia with variable terrestrial ages from 1900 to 24300 years (carbon-14, [2]) were analysed for their halogen distribution from the surface into the interior. The resuits for the chondrites are given in tables 1 to 3. The results show a quite heterogeneous picture. In contrast to samples collected in cold deserts, where a distinct variation with the duration of the exposure to the atmosphere can be recognised, this can easiest be seen for F and I, the concentrations of halogens in chondrites from the Nullarbor plains are mainly altered for Cl and Br. No volume concentrations for H5 chondrites of more than 25 to 50 ppm Cl were found in the Allan Hills. The concentrations of halogens in Reid and Cocklebiddy do not represent original chondrite concentrations. The high contamination in Cl is also reflected in the high Br values. Evident is the quite uniform concentrations of F and I. Methyliodide is no major source for contamination of meteorites from the Plains. We recognise that the diffusion into the interior of a specimen is much more important for meteorites collected in hot environments compared to samples originating from Antarctica. Table 1. Halogens in Reid 006. Depth F Cl Br I [mm] [ppm] [ppm] [ppm] [ppm] 0.0-0.5 15.8 254 1.09 0.42 0.5-1 0 11.0 181 0.59 0.18 1.0-1.5 7.9 241 0.79 0.25 1.5-2.0 8.0 217 0.88 0.19 2.0-3.0 19.0 593 1.50 0.24 3.0-4.0 10.7 491 1.80 0.29 4.0-5.0 10.9 503 1.40 0.30 Table 2: Halogens in Mundrabilla 002. Depth F Cl Br I [mm] [ppm] [ppm] [ppm] [ppm] 0.0-0.5 15.5 191 0.93 0 35 1.0-1.5 14.0 195 0.70 0.25 3.0-4.0 19.0 255 0.91 0.33 5.0-6.0 17.0 219 1.00 0.39 6.0-7.5 18.2 262 1.09 0.36 Table 3: Halogens in Cocklebiddy. Depth F Cl Br I [mm] [ppm] [ppm] [ppm] [ppm] 0.0-0.5 15.3 731 1.9 0.3 2.0-3.0 28 1528 2.8 0.25 5.0-6.0 8.4 1095 2.1 0.19 The measured distribution of the 4 halogens can not be only the result of a single operating process. Beside the contamination a second mechanism, removing some of the elements of interest from outer layers, must have been active. The alteration processes may not only affect the chemical composition but also e.g. determinations terrestrial ages. Is it possible that some of the earlier enrichments of halogens were leached later by liquids? Are there other evidences for the presence of liquid water throughout the terrestrial history of the investigated meteorites? Acknowledgement: All the investigated meteorites were provided by the Museum of Western Australia. References: [1] Langenauer M. and Krahenbuhl U. (1993) Earth Planet. Sci. Lett. 120, 431-442. [2] Jull A.J.T., personal communication. Newton J.* Sephton M. A. Pillinger C. T. Contamination Differences Between CO3 Falls and Antarctic and Saharan Finds: A Carbon and Nitrogen Isotope Study The carbon and nitrogen isotope characteristics of sixteen powdered whole-rock CO3 and related carbonaceous chondrites have been determined using the technique of stepped combustion. The reason for such a study is threefold: as a preview of presolar inventories before more detailed acid residue work, the identification of unusual or even unique specimens, and the evaluation of contamination differences caused by terrestrial weathering. Table 1 shows the list of samples which have been divided into two subgroups: the "normal" CO chondrites which tend to fall into petrographic subtypes 3.2 to 3.6 based on olivine compositions [1] and the primitive CM-CO chondrites which are petrographic subtypes 3.0-3.1. The term CM-CO is used to indicate the similarity between the carbon and nitrogen inventories of these chondrites and members of the CM2 group and particularly the presence in these specimens of presolar silicon carbide: no generic relationship is implied. Acfer 094 possesses mineralogical and chemical characteristics of both groups. It has been described as the first CM3 [2], although on the basis of presolar grain investigations this classification is doubtful [3]. It is better at this stage to refer to it as a unique carbonaceous chondrite. We have shown that there is little correlation between carbon and nitrogen content with petrographic subtype [4], although the CM-COs contain more carbon and nitrogen than the normal COs. Acfer 094 has the highest bulk carbon and nitrogen content of the group. In Table 1 it is apparent that the sample list is a variety of falls and finds which were collected from both cold and hot desert environments. There exists, therefore, an opportunity to study the effects of weathering on carbon and nitrogen content and isotopic composition. Low-temperature carbon released from Antarctic ordinary chondrites is isotopically heavier than non-Antarctic falls ([5] and Fig. 1). We hoped to see a similar situation in the low-temperature carbon in the CO chondrites. All of the meteorites studied here release some carbon below 200 degrees C and the main discussion will be restricted to this material. Nitrogen released at the same temperature is complicated by the presence of heavy nitrogen in the organic matter of the primitive specimens. Although generally the finds contain more nitrogen at 200 degrees C, there are no isotopic differences between the falls and finds. Similarly low-temperature hydrogen is more abundant in CO3 finds [6], but it is isotopically indistinguishable from that in the falls. In contrast, the isotopic composition of low-temperature carbon is related to the circumstances of collection. From Fig. 2 it is evident that low-temperature carbon in the Antarctic finds is isotopically heavier (i.e. enriched in 13C) than the falls, as is the case in ordinary chondrites. Moreover there are other isotopic differences within the sample set. Firstly the Yamato Mountains samples are both isotopically heavier and contain more 200 degrees C carbon than the Allan Hills samples. In contrast low temperature heavy carbon has not been observed in Yamato ordinary chondrites (M.M. Grady pers. comm.). The Saharan samples generally contain heavier carbon than the Allan Hills samples but not quite as heavy as the Yamato Mountains samples. On the basis of Fig. 2 we can argue that the <200 degrees C combustion fraction for samples collected from different environments may be an indicator of the local contamination potential. The nature of this carbon has not been positively identified, but possible sources will be discussed here. All of these samples lose the low temperature heavy carbon when treated with HF/HCl as part of the process of preparing acid-resistant residues. Although this observation is consistent with CO2 induced weathering the combustion temperature rules out mineralogically discrete carbonate. Possible sources of carbon contamination in Antarctica will be discussed first. Carbon dioxide from air has a delta^13C of about -7 per mil and its adsorption onto the Antarctic samples could explain the isotopic composition of most of them. Trapped carbon dioxide in the ice would have a similar composition. One would expect that air adsorption is uniformly active throughout the entire Antarctic meteorite population. However whereas CO3 falls and Antarctic finds are easily distinguishable in terms of low-temperature carbon, the Antarctic and non-Antarctic ordinary chondrites show an overlap in composition (Fig. 1) and in the achondrites this overlap is more pronounced [5]. This behaviour could be caused by the different porosities of these meteorite types which decrease in the order CO3, OC, achondrites. Porosity will affect all types of contamination. Antarctic carbonaceous chondrites are particularly susceptible to surficial evaporite formation, which can form over a relatively short period of time [7]. Bicarbonates such as nesquehonite and hydromagnesite have been analysed in ordinary chondrites, and have been shown to be isotopically heavy (e.g. delta^13C of +7.9 per mil, LEW 85320 nesquehonite, [8]). However it has been tentatively suggested that evaporites forming on carbonaceous chondrites tend to be magnesium sulphates rather than bicarbonates [9]. In addition bicarbonates combust at higher temperatures - 300 to 400 degrees C. However there exists the possibility that the sulphates could contain small amounts of carbon. Most biogenic material tends to be isotopically light -25 to -20 per mil, so cannot be the cause of heavy carbon, although it could affect the falls which have experienced many decades of museum storage. Methyl iodide produced biogenically in the sea and transported by wind could be involved as it is established that the Antarctic meteorites are overabundant in iodine ([10], [11]). However on the basis of the iodine abundance it is difficult to attribute enough carbon to such a source unless the carbonaceous portion of CH3I is accumulated at the expense of the halide following disproportionation. In the hope of recognising any terrestrial organic constituents a sample of ALHA 77003 was subjected to on-line pyrolysis GCMS. A dry-pelleted whole-rock sample was analysed using a pyrojector (SGE Ltd.) at 500 degrees C; GCMS conditions were the same as [12]. Fig. 3 shows the resulting pyrogram. There is no evidence of halogenated compounds which could be equated with seaspray, nor are there any other compounds diagnostic of terrestrial contamination such as isoprenoidal alkanes. Much of the organic matter in ALHA 77003 would appear to be PAHs cleaved from the macromolecule. The delta^13C difference between the two Antarctic collection sites is interesting. One might assume that the nature of the contaminant is the same in both cases, and that the Yamato Mountains samples are simply more weathered than the Allan Hills samples as they contain more low temperature heavy carbon. However if this were the case one would expect the Allan Hills samples to contain the most contamination, as these generally give much older terrestrial ages than the Yamato specimens [13]. Clearly the amount of carbon contamination is not related to terrestrial age. The possible sources of contamination listed above all occur when the meteorites are exposed to air rather than when buried, so a correlation with exposure of the ice surface would be more likely. It is possible that carbon contaminants are controlled by physical geography. In this respect it may be significant that the two collection sites lie on opposite sides of the Antarctic continent. Also the Yamato collection site is nearer to the sea and more exposed than the Allan Hills site. An alternative suggestion is that the difference between the sites is due to different curatorial conditions, as it is well established that discrete generations of salts can grow during storage [8]. The Saharan samples are even more difficult to explain than the Antarctic ones. They could be affected by adsorbed air, biogenic contamination and evaporite formation, but the composition might also be controlled by water passing through the limestone on which they were collected. Terrestrial calcite which has precipitated from dissolved limestone analysed in Algerian meteorites has a delta^13C of +10.0 per mil [14]. The stepped combustion data here suggest the presence of some collection site specific form of contamination or weathering. The understanding of the contribution from various sources is likely to require an in-depth study involving other methods of analysis. References:- [1] Sears D.W.G., Batchelor J.D., Lu J. and Keck B.D. 1991. Proc. Natl. Inst. Pol. Res. 4, 319-343. [2] Bischoff A. and Geiger T. 1994. Lunar Planet. Sci. 25,115-116. [3] Newton J., Bischoff A., Arden J.W., Franchi I.A., Geiger T. and Pillinger C.T. 1994. Meteoritics submitted. [4] Newton J., Arden J.W. and Pillinger C.T. 1992. Lunar Planet. Sci. 23, 985-986. [5] Grady M.M., Wright I.P. and Pillinger C.T. 1991. Geochim. Cosmochim. Acta 55, 49-58. [6] Morse A.D., Newton J. and Pillinger C.T. 1993. Lunar Planet. Sci. 24, 1017-1018. [7] Jull A.J.T., Cheng S., Gooding J.L. and Velbel M.A. 1988. Science 243, 417-419. [8] Grady M.M., Gibson E.K. Jr., Wright I.P. and Pillinger C.T. 1989. Meteoritics 24, 1-7. [9] Velbel M.A. 1988. Meteoritics 23, 151-159. [10] Heumann K.G., Gall M. and Weiss H. 1987. Geochim. Cosmochim. Acta 51, 2541-2547. [11] Langenauer M. and Krahenbuhl U. 1993. Earth Planet. Sci. Lett. 120, 431-442. [12] Gilmour I. and Pillinger C.T. 1994. Mon. Not. R. Astron. Soc. 269, 235-240. [13] Nishiizumi K., Elmore D. and Kubik P.W. 1989. Earth Planet. Sci. Lett. 93, 299-313. [14] Grady M.M. and Pillinger C.T. 1993. Earth Planet. Sci. Lett. 116, 165-180. Meteorite Fall/find Location Pet. type Group Acfer 094 Find Algeria 3.0 Unique ALHA 77307 Find Antarctica 3.1 Colony Find Oklahoma 3.0 "CM-CO" Y 81020 Find Antarctica 3.3 Acfer 202 Find Algeria 3.5 Acfer 243 Find Algeria 3.7 ALH 82101 Find Antarctica 3.4 ALHA 77003 Find Antarctica 3.4 Felix Fall Alabama 3.4 Kainsaz Fall Russia 3.2 "normal" CO Lance Fall France 3.4 Ornans Fall France 3.4 Warrenton Fall Missouri 3.6 Y 791717 Find Antarctica 3.3 Y 82050 Find Antarctica 3.3 Y 82094 Find Antarctica 3.5 Table 1. Sample list with collection details of each specimen, petrographic type (Sears et al. 1991) and subgroups. Overleaf:- Fig. 1. delta^13C of the 200 degrees C carbon step in ordinary chondrite (Grady et al. 1991) and CO3 chondrite falls and Antarctic finds. Fig. 2. delta^13C of the carbon released below 200 degrees C in the CO chondrites. Fig. 3. ALHA 77003 pyrogram. Scherer P.* Schultz L. Loeken T. Weathering and Atmospheric Noble Gases in Chondrites from Hot Deserts Since the last decade a large number of new meteorites have become available from hot desert regions mainly from Africa, Australia and America. Most of these samples are more severely weathered than the relatively fresh specimens found on blue icefields in Antarctica or modern falls. To determine the weathering grade in the thin section and to investigate the most common alteration products was one aim of this project. Another important purpose of this work was to measure and to compare isotopic composition and concentrations of all noble gases in ordinary chondrites from hot deserts with those from other areas and to determine diverging noble gas patterns due to their alteration effects on earth. There are three major trapped noble gas components found in ordinary chondrites: the intrinsic planetary gases, solar gases and terrestrial atmospheric gases as "contamination". These three major components can be separated from each other as well as from in situ produced cosmogenic or radiogenic gases by the use of their different elemental and isotopic compositions. Apart from the isotopic composition of Xe the contribution of atmospheric noble gases is also visible in the concentrations of the heavy noble gases Kr and Xe. As first shown by Zahringer (1966) and Marti (1967) the petrographic-chemical type of chondrites is correlated with the concentrations of the planetary 84Kr and 132Xe, e.g. type 3 is characterized by 84Kr > 8 10^(-10) cm^3 STP/g and type 5 and 6 by 84Kr < 1 10^(-10) cm^3 STP/g. The correlation line shown in Fig. 1 is calculated as a best linear fit through data points of H3 and H4 chondrite falls (data from Schultz et al.,1990). Almost all chondrites from hot desert regions plot to the right of this correlation line indicating the presence of additional Kr and, to a lesser extent, also Xe. The fact that these samples fall below rather than above the correlation line for uncontaminated specimens is due to the atmospheric 84Kr/132Xe ratio of about 28 which is much higher than the planetary value of about 0.6. Acfer 074 (L6), which lies far outside the range of Fig.1, shows the highest measured 84Kr concentration. With more than 1*10^(-8) cm^3 STP/g it contains about 100 times more 84Kr than the Bruderheim (L6) standard. The concentrations of 84Kr and 132Xe in bulk samples of chondrites from hot desert regions are not consistent within one specimen (Fig. 2). For the Antarctic meteorite ALH 88007 the variation from the mean in the 84Kr concentration is about 13%. In Acfer 019, however, 84Kr varies by about a factor of two, or from the lowest to the highest value by more than a factor of six. The elemental ratios of Ar, Kr, and Xe provide information on the mechanisms of gas incorporation into meteoritic matter. The measured meteorites from hot deserts are plotted together with H-chondrite falls and Antarctic H-chondrites (data from SCHULTZ et al.,1991) in Fig.3. This plot reveals that the high Kr concentrations in hot desert meteorites are roughly correlated with low 132Xe/ 84Kr ratios. Most data points are arranged in three clusters. The H-chondrite falls (I) as the least weathered samples plot around the average value for planetary gases in ordinary chondrites with 132Xe/ 84Kr ~ 1.7. The slightly to moderately weathered Antarctic meteorites scatter between cluster I (falls) and cluster III (measured hot desert meteorites), where most data points intercept the y-axis at 132Xe/84Kr ~ 0.15. This ratio is close to the atmospheric value of 0.036 and the fact that the hot desert meteorites never reach this point implies that an elemental fractionation takes place during the process of trapping caused by a higher adsorption coefficient for Xe than for Kr (e.g.Ozima and Podosek,1983). Another explanation is that heavy noble gases are fractionated due to solution in water before incorporation in secondary weathering minerals (e.g. iron hydroxides or evaporates) that started to built up in the meteorites during their exposure to the surrounding soil and atmosphere on earth. The latter explanation is supported amongst other things by the fact that the 132Xe/ 84Kr ratio of 0.15, which is found in severely weathered chondrites from hot deserts, is close to the value of 0.073 determined for water at 0 degrees C (Ozima and Podosek,1983). The results concerning the bulk composition of noble gases in chondrites from hot deserts have shown that adsorption in connection with elemental fractionation, possibly together with in water dissolved noble gases, are responsible for the effect of terrestrial contamination in chondrites. The host phases are possibly iron oxides, iron hydroxides and evaporates. This is also seen in the trend that the amount of adsorbed heavy noble gases correlates with the grade of alteration determined in the thin section (weathering classification after Wlotzka, 1993). Several experiments, like the treatment of some hot desert meteorites with oxalic acid or HCl to remove selectively terrestrial alteration products (rust or carbonates), were carried through to get considerable information on the host phases for trapped atmospheric noble gases in chondrites. As a result most of the acid residues show lowered concentrations of atmospheric Kr and Xe and the Xe isotopic ratios are shifted towards the AVCC composition. The fact that the chemical treatment does not completely remove the atmospheric gases in these samples reveals that there appear to be other additional mechanisms that are responsible for trapping terrestrial noble gases. Stepwise heating experiments of some samples show that the atmospheric gases cannot be completely removed below temperatures of 1200 degrees C, thus, some of the trapped gases are tightly bound. This excludes simple physical adsorption as an explanation for the process which introduces the gases to meteoritic material because preheating the samples at 140 degrees C for at least 48 hours has not removed this component. Niedermann and Eugster (1992) describe a process called "irreversible adsorption" for noble gas trapping on mineral surfaces during their crushing in spiked atmospheres. This "anomalous adsorption" probably takes place also in other natural mechanical weathering processes initiated amongst other things by large day and night temperature changes in the hot desert regions. However, the incorporation of noble gases in newly formed secondary weathering products and/or trapping in cavities caused by mechanical and chemical treatment in the laboratory are likely to be important processes which change the original noble gas composition in meteorites. References: Marti K. (1967) Earth Planet. Sci. Lett., 2, 193-196. Niedermann S. and Eugster O. (1992) Geochim. Cosmochim. Acta, 56, 493-509. Ozima M. and Podosek F.A. (1983) Noble gas geochemistry. Cambridge University Press, pp. 367. Schultz L., Weber H.W., and Begemann F. (1990) Meteoritics, 25, 405-406. Schultz L., Weber H.W., and Begemann F. (1991) Geochim. Cosmochim. Acta, 55, 59-66. Wlotzka F. (1993) Meteoritics, 28, 460. Zahringer J. (1966) Earth Planet. Sci. Lett., 1, 379-382. Fig.1: The concentration of planetary 84Kr and 132Xe in ordinary chondrite falls is correlated with their petrographic-chemical type. The correlation line describes the best linear fit through H3 and H4 falls (data from SCHULTZ et al.,1990). The bar in the "type 4" section indicates the range of scattering for the used data points. The investigated hot desert meteorites, however, do not plot along this line. They drift to the right, which is due to an additional contribution of atmospheric gases with a higher Kr/Xe ratio compared to the planetary value. The dotted line indicates, as an example, the replacement from a possible source region in the "type 6" field to the actually measured data point. Fig. 2: Variation of the 84Kr concentration in the H5-chondrite ALH 88007 (wide bars) and the hot desert meteorite Acfer 019 (L6) for eight individual measurements. Fig. 3: Three major clusters can be identified in this plot: (I) relatively "fresh" falls, (II) slightly alterated Antarctic samples and (III) the more severely weathered meteorites from hot desert regions. The data points for these highly alterated samples drift down to the atmospheric ratio of 0.036. Elemental fractionation is the reason for not reaching this value for the terrestrial atmosphere [A]. The y-axis intercept of about 0.15 for some of the hot desert meteorites lies closer to the 132Xe/84Kr ratio of 0.073 determined for water [W] at 0 degrees C. Thursday, July 21, 1994 COSMOGENIC NUCLIDES 11:20 - 12:30 p.m. Chair(s): D. W. G. Sears Knauer M.* Neupert U. Michel R. Bonani G. Dittrich-Hannen B. Hajdas I. Ivy Ochs S. Kubik P. W. Suter M. Measurement of the Long-Lived Radionuclides 10Be, 14C, 26Al in Meteorites from Hot and Cold Deserts by Accelerator Mass Spectrometry (AMS) Introduction: Cosmogenic nuclides provide a unique record of collision, exposure and terrestrial histories of meteorites. The determination of terrestrial ages is of particular importance for meteorites which are found in large numbers in hot and cold deserts. In the course of a project to study meteorites from hot and cold deserts, we investigate the nuclides 10Be, 14C and 26Al in meteorites from Antarctica and from the Sahara. Here we report first results and discuss some aspects of their interpretation on the basis of production rates derived from physical model calculations. Experimental: Procedures for the separation of 10Be, 14C, and 26Al from stony meteorites and for the preparation of AMS samples were established at Hannover for the first time. The AMS measurements are made at the ETH/PSI AMS laboratory at the ETH Honggerberg in Zurich. A separation scheme of Vogt and Herpers (1988) for 10Be and 26Al was modified in order to avoid use of perchloric acid and to improve purities and yields of the BeO and Al2O3 products for the AMS measurement (Knauer, 1994). After adding 2 mg Be carrier and 2 mg Al carrier to the crushed meteorite material, pressure digestion by a mixture of nitric acid, hydrochloric acid and sulphuric acid is performed. After evaporating the solution and pressure digestion of the residue with hydrochloric acid, an aliquot is taken for determination of the total Al by ICP-OES. After this, Fe is extracted by methylisobutylketone. Be and Al are separated by cation exchange (Dowex 50 WX8) and precipitated as hydroxides, which are glowed in quarz crucibles to oxides. The oxides are mixed with copper for AMS measurement. All steps of the separation scheme were investigated in detail by ICP-OES before meteorites were analyzed. The quality of the 10Be and 26Al determination was checked by analyzing a number of meteorite falls (Knauer, 1994). The results are in good agreement with literature data. In order to investigate the precision of the analyses, a meteoritic standard was prepared from the LL6 chon- drite Dhurmsala. 74 g of Dhurmsala, split 2/3a, were crushed to a grain size of less than 125 micrometers. Metal grains (about 1 g) which could not be crushed to this size were removed. The resulting material now serves as an interlaboratory standard for the laboratories at Koln and Hannover. Repeated analyses of this standard yielded detailed information on the uncertainties of the 10Be and 26Al determination (table 1). For the determination of 14C, carbon is extracted from meteorite samples by high frequency heating and transformed to CO2. Meteorite samples of about 150 mg are placed in single use Al2O3 crucibles between two high- purity tungsten rods for proper coupling of the high frequency. In a first step, a sample is heated for 30 min to 1000 degrees C to remove terrestrial contamination and weathering products. Then a first fraction of the gas is taken as a control for AMS measurement. Afterwards the sample is kept melting for about 2 min at 1700 degrees C. Then the temperature is lowered to 800 degrees C and this temperature is maintained for about 20 min. This second fraction is used to determine the cosmogenic 14C. Both fractions are completely oxidized, the CO2 is purified and diluted with about 6 mg 14C-free CO2. Further details of the procedure may be found elsewhere (Knauer, 1994). The reduction of CO2 to C and the subsequent AMS measurement are made in Zurich. Details on the reduction process and on the AMS measurements of 14C can be found elsewhere (Bonani et al., 1987; Vogel et al., 1984, 1987). The 14C analyses were checked by analyzing samples of the Bruderheim chondrite. Specific activities of 9.7 +/- 0.4 dpm/kg and 47.6 +/- 2.0 dpm/kg were found in the carbon fractions taken below and above 1000 degrees C, respectively. According to pyrolysis experiments on some ordinary chondrites from the Acfer region (Sahara), performed at the Friedrich Schiller Universitat Jena, the fraction above 1000 degrees can be interpreted as carrying the cosmogenic 14C. The value of 47.6 +/- 2.0 dpm/kg lies right inbetween the AMS results reported by Jull et al. (1989) and by Beukens et al. (1988) (table 1). A 14C analysis of Dhurmsala, split 2/3a, resulted in 53.0 +/- 2.3 dpm/kg for the fraction above 1000 degrees C. Both our results for Bruderheim and Dhurmsala lie in the range of 14C concentrations observed in stony meteorite falls, i.e. between 38 dpm/kg and 60 dpm/kg as reported by Jull et al. (1989). Table 1: Contribution of chemistry and AMS errors to the uncertainty of the 10Be and 26Al determination (1s) as determined from n analyses of the Dhurmsala standard and comparison of 14C analyses of Bruderheim from Beukens et al. (1988), Jull et al. (1989) and from this work. average n experimental error [%] [dpm/kg] chemistry AMS total 10Be 21.4 +/- 1.1 18 4.7 2.1 5.1 26Al 69.7 +/- 3.3 9 3.3 3.4 4.7 Meteorite 14C[dpm/kg] reference Bruderheim 46.8 +/- 1.4 Jull+ (1989) Bruderheim 47.6 +/ 2.0 this work Bruderheim 50.1 +/- 0.3 Beukens+ (1988) Model Calculations: Earlier calculations of depth and size dependent GCR production rates by a physical model (Michel et al., 1991; Bhandari et al., 1993) are extended with respect to coverage of cosmogenic nuclides, meteoroid classes and sizes (Michel et al., 1994; Herpers et al., 1994). In this work, we show production rate depth profiles in H-chondrites of 10Be and 26Al (Fig. 1) and 14C (Fig. 2). In addition, we present correlations between 10Be and 26Al production rates (Fig. 1), between 14C/10Be and 26Al/10Be ratios (Fig. 2) and the dependence of 14C on 3He/21Ne and 22Ne/21Ne ratios (Fig. 3). The calculations of 10Be, 26Al and of helium and neon isotopes are consistent with recent calculations (Michel et al., 1994; Herpers et al., 1994). They represent an improved status as a consequence of the analysis of results from a simulation exper- iment in which an artificial meteoroid made of gabbro (R = 25 cm) was irradiated isotropically with 1600 MeV protons. The calculations of 14C have still lower quality than those of the other isotopes, because of a considerable lack of thin-target production cross sections for proton- induced reactions. Moreover, a validation of the model calculations by analyzing results from simulation experiments is not yet available. Allan Hills Ordinary Chondrites - Results and Discussion: 10Be and 26Al were determined in 30 Allan Hills ordinary chondrites from the 1988 campaign (table 2). Since rare gas measurements are not yet available except for ALH 88019, a final discus-sion will be postponed until these measurements which are planned for the near future at Mainz are made. For 16 of these meteorites measurements of natural thermoluminescence (NTL) were reported by Benoit et al. (1991) who proposed pairing of H5: 88025, 88038; H5: 88014, 88040; H5: 88026, 88030, 88033; H5: 88029, 88042. From the results of our analyses there are no objections against a pairing of ALH 88014 and ALH 88040. The lack of correlation between 10Be and 26Al makes it hard to accept the other three pairings. A final discussion will be possible when the rare gas data are available. Table 2: Cosmogenic radionuclides in 30 Allan Hills ordinary chondrites from the 1988 campaign. For some of the meteorites measurements of natural thermoluminescence (NTL) at 300 degrees C exist from Benoit et al. (1991). Meteorite 10Be 26Al NTL (dpm/kg) (dpm/kg) (krad) ALH88011 H3 14.8+/-0.9 38.3+/-2.9 ALH88020 H3 18.3+/-1.2 42.8+/-2.5 210.0+/-2. ALH88036 H3 12.0+/-0.9 42.3+/-2.5 20.0+/-9. ALH88013 H4 15.1+/-0.9 52.6+/-3.9 ALH88016 H4 19.9+/-1.2 56.4+/-4.1 ALH88017 H4 20.0+/-1.2 56.8+/-4.8 ALH88008 H4/5 18.6+/-1.1 55.6+/-4.3 ALH88010 H4/5 16.6+/-1.0 42.7+/-3.1 ALH88031 H4/5 17.0+/-0.9 56.1+/-3.2 ALH88014 H5 15.5+/-1.0 56.5+/-4.1 32.0+/-0.1 ALH88019 H5 5.6+/-0.4 10.4+/-1.2 ALH88025 H5 16.1+/-0.9 52.1+/-3.2 27.7+/-0.3 ALH88026 H5 20.8+/-1.1 56.2+/-3.4 127.1+/-0.5 ALH88027 H5 19.4+/-1.1 55.4+/-3.7 176.0+/-1. ALH88028 H5 15.4+/-1.0 37.0+/-2.3 0.8+/-0.1 ALH88029 H5 18.1+/-1.1 53.4+/-4.4 226.0+/-2.0 ALH88030 H5 18.5+/-1.0 48.4+/-3.1 120.0+/-3.0 ALH88032 H5 15.1+/-0.8 34.7+/-3.0 60.0+/-0.4 ALH88033 H5 17.0+K/-0.9 62.3+/-3.4 117.5+/-0.5 ALH88038 H5 15.1+/-1.1 61.5+/-3.6 27.0+/-0.3 ALH88039 H5 16.6+/-1.1 58.1+/-3.5 145.1+/-0.1 ALH88040 H5 17.0+/-0.9 60.8+/-3.1 32.0+/-0.1 ALH88042 H5 15.7+/-0.8 58.6+/-3.1 238.3+/-0.3 ALH88018 H6 19.4+/-1.2 53.1+/-3.5 ALH88021 H6 17.5+/-0.9 51.9+/-3.3 ALH88043 H6 17.4+.-0.9 58.6+/-2.8 ALH88002 L4 18.6+/-1.1 52.1+/-3.4 ALH88012 L6 18.1+/-1.2 65.3+/-5.4 ALH88024 L6 16.4+/-0.9 36.5+/-2.3 9.4+/-0.1 ALH88004 LL4 13.2+/-0.8 73.4=/-4.5 With the exception of ALH 88019 all 26Al data are in the range observed in H- and L- chondrite falls or of production rates calculated for meteoroids with radii up to 50 cm (Fig. 1), though some of them are probably affected by substantial terrestrial ages. 10Be data of four meteorites (ALH88004, ALH88011, ALH88019, and ALH88036) are too low to fit into the range of 10Be activities in normal-sized meteorites (Fig. 1). However, a mere comparison of individual radionuclide activities with possible produc- tion rate ranges is not sufficient. The correlation between production rates of cosmogenic radio- nuclides must also be taken into account. Such a correlation is shown in Fig 1, defining a field of allowed 10Be and 26Al concentrations, provided that the activities of these meteorites are in satu- ration. For meteoroid radii between 5 cm and 40 cm, there is a nearly linear correlation between 10Be and 26Al production rates. For larger radii the linear correlation breaks down as a conse- quence of 10Be being a medium-energy product and 26Al being a low-energy one. If 10Be and 26Al are in saturation, their correlation can give some evidence of size of the meteoroid and of shielding depth for meteoroid radii above 40 cm. For smaller radii the information on size and shielding becomes more ambiguous. Comparing the experimental results with the allowed field of 10Be-26Al combinations makes it possible to identify long terrestial ages, complex exposure histories and possible contri- butions of solar cosmic ray interactions. A comparison of the 10Be and 26Al results of the Allan Hills chondrites shows that most results can be interpreted as normal depth and size effects, though some 10Be values are relatively low. The data of ALH88011 and ALH88036 could be explained by these meteorite samples coming from large meteoroids. A final discussion of this point, which has also to take into account the possibility of old terrestrial ages, can only be made when the rare gas data are available. Fig. 1: Calculated production rate depth profiles of 10Be and 26Al in H- chondrites (meteoroid radii from 5 cm to 120cm and for a 2p irradiation geometry) and correlation between calculated 26Al and 10Be production rates (H-chondrites with radii between 5 cm and 120 cm). The results for ALH88004 do not fall into the allowed field of 10Be-26Al data explainable by GCR interactions. This meteorite had a recovered mass of 315.7 g which gives a lower limit of its preatmospheric radius (2.8 cm). A possible explanation could be that this meteorite comes from a very small (R < 5 cm) meteoroid and that the high 26Al activity is a result of SCR interac- tions. Also this question can only be answered when rare gas data are available. In the case of ALH88019, the situation is more complicated. Scherer (1993) measured He, Ne and Ar-isotopes and derived a mean exposure age of 41.26 Ma based on T(exp,3) = 33.62 Ma, T(exp,21) = 42.03 Ma and T(exp,38) = 40.48 Ma. He assumed a 3He loss of 18.5 % though the gas retention ages are 4.16 Gy and 3.5 Gy for for 40Ar and 4He, repectively. According to the rare gas data, 10Be and 26Al were in saturation at time of fall if ALH 88019 had a single stage exposure history. The measured 10Be (5.6 +/- 0.4 dpm/kg) and 26Al (10.4 +/- 1.2 dpm/kg) data are extremely low and fall sigificantly outside of the allowed field of 10Be/26Al values. These extraordinary results were confirmed by repeated analysis for the same sample and by analysis of a new sample which was analyzed before by Scherer (1993) for rare gases. These analyses resulted in 6.0 +/- 0.4 dpm/kg, 5.4 +/- 0.5 dpm/kg, and 5.4 +/- 0.5 dpm/kg for 10Be and 10.3 +/- 1.2 dpm/kg and 10.5 +/- 1.3 dpm/kg for 26Al. A third 26Al sample awaits for AMS measurement. A detailed discussion of the results obtained for ALH88019 will be performed when this measurement and planned investiga- tions of cosmic ray tracks are finished. Ordinary Chondrites from Sahara - Results and Discussion: Thirteen ordinary chon- drites from Sahara were analyzed for 10Be, 14C and 26Al (table 3). For three of these meteorites, 26Al has still to be measured. Rare gas measurements do not yet exist. There are some 10Be and 26Al data which fall outside the systematics. Extremely low 10Be concentrations are found in Hammadah al Hamra 004 and Ilafegh 013. For these two meteorites the 26Al concentrations are also low. The data for Acfer 129 can only be explained as production rates. The high 10Be value of 21 dpm/kg proves 26Al and 14C to be production rates. 26Al of 88.7 dpm/kg is exceptionally high and our only explanation is that we have some 26Al produced by solar cosmic rays in this meteorite. All 14C concentrations measured in the 13 ordinary chondrites from the Sahara are sub- stantially lower than for falls. A comparison of the measured 10Be activities and the 10Be produc- tion rate ranges makes sure that 14C was in saturation in all meteorites at time of fall. In order to analyze these data with respect to terrestrial ages, 14C production rates have to be assumed. According to our model calculations the 14C GCR production rates in chondrites vary between 18. and 57. dpm/kg in H-chondrites (Fig. 2) and between 19. and 60. dpm/kg in L- chondrites. Therefore, depth- and size-corrected production rates should be used for the determination of terrestrial ages. Such production rates can be obtained by model calculations from correlations between 14C production rates and production rates of other radioactive and/or stable cosmogenic nuclides. Table 3: Analyses of ordinary chondrites from the Sahara: cosmogenic radionuclides (10Be, 14C, and 26Al), terrestrial ages and natural thermoluminescence (NTL) at 300 degrees C. NTL data are from Benoit et al. (1993). The 14C specific activities were derived from the extraction above 1000 degrees C. There are two entries for terrestrial ages: T(ter) are ages based on depth- and size corrected 14C production rates (P14), T(ter,44) denotes terrestrial ages based on an average 14C production rate of 44 dpm/kg. * The maximum production rate P14 was assumed since 26Al is probably affected by SCR produced 26Al. $ samples not yet measured. METEORITE CLASS 10Be 26Al 14C P(14) (dpm/kg) (dpm/kg) (dpm/kg) (dpm/kg) Acfer 023 H3 20.4+/-1.5 61.4+/-4.7 3.5+/-0.2 47.9 21.6 Acfer 129 H3 21.0+/-1.4 88.7+/-6.4 4.4+/-0.2 56.7* 21.1 Acfer 153 H3 15.6+/-1.3 $ 8.5+/-0.4 $ Acfer 171 H3.7 18.7+/-1.3 $ 15.2+/-0.8 $ Acfer 022 H3.7 17.4+/-1.1 53.7+/-4.2 13.5+/-0.6 42.1 Acfer 028 H3.8 20.4+/-1.5 53.2+/-4.5 14.7+/-0.6 40.8 Acfer 039 L3 17.6+/-1.2 46.3+/-4.1 18.1+/-0.8 35.2 Acfer 066 L(LL)3 17.6+/-1.2 51.9+/-5.0 5.9+/-0.3 39.1 Acfer 080 L3.9 14.5+/-1.1 40.6+/-3.4 18.0+/-0.8 30.7 Adrar 003 LL(L)3 23.3+/-1.2 $ 19.9+/-1.0 $ Ham.Ham.004 H3 13.0+/-0.9 28.9+/-2.3 10.5+/-0.5 21.5 Ilafegh 013 H3 9.1+/-0.7 31.1+/-2.5 22.2+/-1.1 25.2 Tanezr. 006 H3 16.2+/-1.1 47.0+/-3.9 13.2+/-0.7 36.5 METEORITE CLASS T(ter) T(ter,44) NTL (ka) (ka) (krad) Acfer 023 H3 21.6 20.9 3.1+/-1.2 Acfer 129 H3 21.1 19.0 0.44/-0.02 Acfer 153 H3 $ 13.6 7.2+/- 0.1 Acfer 171 H3.7 $ 8.8 20.5+/-5.0 Acfer 022 H3.7 9.4 9.8 0.10+/-0.01 Acfer 028 H3.8 8.4 9.1 0.43+/-0.08 Acfer 039 L3 5.5 7.3 9.0+/-1.0 Acfer 066 L(LL)3 15.6 16.6 12.0+/-1.0 Acfer 080 L3.9 4.4 7.4 28.0+/-3.0 Adrar 003 LL(L)3 $ 6.6 Ham.Ham. 004 H3 5.9 11.8 8.0+/-3.0 Ilafegh 013 H3 1.0 5.7 23.0+/-2.0 Tanezr. 006 H3 8.4 10.0 4.0+/-2.0 In Fig. 2, calculated production rates of 14C, 10Be and 26Al are presented in the form of a three-isotope plot, plotting 14C/10Be versus 26Al/10Be. There is a nearly linear correlation between these production rate ratios for meteoroid radii between 5 cm and 120 cm. Both ratios increase with depth inside a given meteoroid. Further the ratios in the center of meteoroids increase with radius. The correlation is ambiguous since the ranges of production rate ratios overlap for diffe- rent meteoroid radii. But it allows to derive depth and size corrected terrestrial ages if same- sample measurements of 10Be, 14C, and 26Al exist. One can derive from it the proper production rate of 14C, provided 10Be and 26Al are in saturation. Then we can calculate from the production rates depth and size corrected terrestrial ages. This procedure needs a rare gas measurement of the meteorite to assure saturation, but this measurement must not necessarily be a same-sample measurement. If, however, a same-sample measurement of rare gases is available, the proper 14C production rate can also derived from its correlation with production rate ratios such as 22Ne/21Ne or 3He/21Ne ratios (Fig. 3). Though we do not yet have the rare gas measurements, which could assure that 10Be and 26Al are in saturation, the new 10Be, 14C and 26Al data of the Sahara meteorites were interpreted exemplarily as if 26Al and 10Be were in saturation (table 3). A comparison of terrestrial ages, based on the shielding corrected production rates, Tter, and those derived by assuming a mean production rate of 44 dpm/kg, T(ter,44), clearly demonstrates, that shielding corrections should not be neglected for terrestrial age determination. Moreover, the observation of production rates in excess of 44 dpm/kg in stony meteorite falls limits the applicability of the method. The terrestrial ages vary between 1 ka and 21.1 ka with three meteorites having substantial terrestrial ages of 21.1, 21.6 and 15.6 ka, while the others are all below 10 ka. Fig. 2: Depth profiles of calculated 14C production rates in H- chondrites (meteoroid radii from 5 cm to 120 cm andfor a 2pi irradiation geometry) and three isotope plot of calculated production rates, 14C/10Be versus 26Al/10Be,(H-chondrites with radii between 5 cm and 120 cm). With the help of these data some remarks considering pairing can be made. We cannot exclude pairing of Acfer 129 and Acfer 023 having the same terrestrial age. Both have similar 10Be, but slightly different 26Al and different 14C. The differences in 26Al could be the result of Acfer 129 coming from the surface and Acfer 023 from the interior of the same meteoroid. This would require that the lower NTL of Acfer 129 is caused by heating of the near-surface parts of the meteoroid. This matter can only be settled after the rare gas data for the other meteorites for which all these radionuclides were measured become available. Combining the terrestrial ages, 10Be and 26Al and the NTL data, for the other meteorites there are no indications of pairing and it can be proposed that these meteorites are from different falls. For meteorites from hot deserts with their moderate terrestrial ages, 14C is a good tool to distinguish different falls and to exclude pairing. If we use shielding corrected terrestrial ages, the time resolution should be sufficient to exclude pairing for these meteorites. Fig. 3: Dependence of 14C production rates in H-chondrites on 22Ne/21Ne and 3He/21Ne production rate ratios formeteoroid radii between 5 cm and 120 cm. Acknowlegements: The Allan Hills meteorite samples were provided by EUROMET, the Sahara meteorites were obtained from the Institut fur Planetologie/Universitaet Munster, the Bruderheim sample from MPI Chemie/Mainz and the Dhurmsala material from Abteilung Nuklearchemie/Universitaet zu Koln. The authors thank Prof. Dr. K. Heide/Friedrich Schiller Universitat Jena for the investigation of the thermal outgassing of stony meteorites. This work was supported by the Deutsche Forschungsgemeinschaft. References: Benoit P.H., Sears H., Sears D.W.G. (1991) Meteoritics 26, 262 Benoit P.H., Jull A.J.T., McKeever S.W.S., Sears D.W.G. (1993) Meteoritics 28, 196 - 203 Beukens R.P., Kucklidge J.C., Miura Y. (1988) Proc. NIPR Symp. Antarct. Meteorites 1, 224 - 230 Bhandari N., Mathew K.J., Rao M.N., Herpers U., Bremer K., Vogt S., Wolfli W., Hofmann H.-J., Michel R., Bodemann R., Lange H.-J. (1993) Geochim. Cosmochim. Acta 57, 2361-2375 Bonani G., Beer J., Hofmann H.-J., Synal H.-A., Suter M., Wolfli W., Pfleiderer Ch., Kromer B., Junghans Ch., Munnich K.-O. (1987) Nucl. Instr. Meth. Phys. Res. B29, 87 - 90 Eberhardt P., Eugster O., Geiss J., Marti K. (1966) Z. Naturforschg. 21a, 414 - 426, 1966 Herpers U., Vogt S., Bremer K., Hofmann H.J., Suter M., Wolfli W., Wieler R., Lange H.-J., Michel R. (1994) Planetary Space Science, in press Jull A.J.T., Donahue D.J., Linick T.W. (1989) Geochim. et Cosmochim. Acta 53, 2095 - 2100 Jull A.J.T., Wlotzka F., Bevan A.W.R., Brown S.T., Donahue D.J. (1993) Meteoritics 28, 376 - 377 Knauer M. (1994) thesis, Universitat Hannover Michel R., Dragovitsch P., Cloth P., Dagge G., Filges D. (1991) Meteoritics 26, 221 - 242 Michel R., Lange H.-J., Lupke M., Herpers U., Rosel R., Suter M., Dittrich-Hannen B., Kubik P.W., Filges D., Cloth P., (1994) Planetary and Space Science, in press Scherer P. (1993) thesis, Universitat Mainz Vogt S., Herpers U. (1988) Fresenius Zeitschr. Anal. Chem. 331, 186 - 188 Vogel J.S., Southon J.R., Nelson D.E. (1984) Nucl. Instr. Meth. Phys. Res. B29, 289 - 293 Vogel J.S., Southon J.R., Nelson D.E. (1987) Nucl. Instr. Meth. Phys. Res. B29, 50 - 56 Meltzow B.* Herpers U. Dittrich-Hannen B. Kubik P. W. Suter M. 10Be and 26Al Concentrations in Carbonaceous Chondrites and Other Meteorite Types from the Sahara Introduction Stable and long-lived products of cosmic ray interactions with extraterrestrial matter constitute a unique record of the history of small bodies in the solar system. Measurements of these cosmogenic nuclides in meteorites allow the investigation of collision and exposure histories of the parent bodies. The measurements of 10Be and 26Al provide additional clues to possible pairings. Experimental The investigation reported herein encloses 22 samples from the Algerian Sahara and one from Mongolia, kindly placed at our disposal by the Institut fur Planetologie, Munster. 10Be and 26Al concentrations were measured by means of accelerator mass spectrometry (AMS). Samples for this method were prepared using a chemical procedure which has been described in detail previously [1,2]. The masses of the samples prepared for AMS measurements varied from 50 mg to 250 mg. The AMS measurements were performed at the PSI-ETH AMS facility in Zurich. Descriptions of the AMS technique are given in [3,4]. For 10Be, the standard "S555" with a nominal 10Be/9Be ratio of 95.5 x 10^(-12) was used, for 26Al, the standards used were "Al9" and "AL1092" with a nominal Al / 27Al ratio of 1190 x 10^(-12) and 134 x 10^(-12), respectively. The half-lives utilized to convert the measured data of 10Be and 26Al nuclides into activities are 1.51 Ma and 716 ka for 10Be and 26Al respectively. 26Al was also measured nondestructively by gamma-gamma-coincidence counting. The technique used has been described in detail elsewhere [5,6]. Samples investigated in this manner had masses from 3 g up to 37 g. The accuracy of this method is about 5 %. In Fig. 1 the correlation between the 26Al concentrations determined by gamma-gamma-coincidence counting and AMS is shown. The individuel measurements are represented by different symbols, the solid line was obtained by linear regression. The results of both methods agree, except for a few data points, within the range of the errors. Discrepancies between both methods are not necessarily erroneous, but can rather be explained by the different sizes of the samples used. Due to depth effects, SCR effects or chemical inhomogeneities, the activity of a cosmogenic nuclide is not uniform within a large sample. Figure 1, showing a comparison of the 26Al concentrations measured by means of AMS and gamma-gamma-coincidence spectrometry, appears here in the hard copy. Results and discussion The results of the 10Be and 26Al determinations in C chondrites are given in table 1, the results of mesosiderites and other classes of Saharian meteorites are given in table 2. The classification of the meteorites indicated in the tables was done by Bischoff et al. [7]. For some of the meteorites, exposure ages based on 21Ne determinations [8- 10] are available. These ages are also included in the tables. Table 1. 10Be and 26Al concentrations in C chondrites from the Sahara measured by means of AMS and gamma-gamma-coincidence-spectrometry Meteorite Sample Class 26Al [dpm/kg] gamma-gamma AMS Acfer 094 A 094 CM3 36.8+/-1.8 33.2 +/- 2.1 Acfer 202 A 202 CO3 57.3 +/- 2.9 49.4 +/- 8.4 Acfer 243 A 243 CO3 28.7 +/- 1.7 40.2 +/- 1.8 40.2 +/- 2.4 Acfer 087 A 87 CR 42.2 +/- 1.7 36.9 +/- 3.9 Acfer 186 A 186 CR 46.8 +/- 1.9 40.9 +/- 3.2 El Djouf 001 Dy 1 CR 48.7 +/- 2.4 Acfer 082 A 82 CV3 26.0 +/- 1.3 27.7 +/- 1.6 Acfer 086 A 86 CV3 35.5 +/- 1.7 32.9 +/- 1.6 Acfer 272 A 272 CV3 42.3 +/- 2.2 37.7 +/- 3.1 Acfer 182 A 182 CH 29.8 +/- 1.2 30.8 +/- 2.4 Acfer 207 CH 29.5 +/- 1.5 Meteorite Sample Class 10Be [dpm/kg] T(rad) AMS [Ma] Acfer 094 A 094 CM3 14.6 +/- 0.3 Acfer 202 A 202 CO3 3.5 +/- 0.8 Acfer 243 A 243 CO3 14.4 +/- 0.3 15.0 +/- 0.3 Acfer 087 A 87 CR 17.5 +/- 0.7 6 Acfer 186 A 186 CR 21.5 +/- 0.3 6 El Djouf 001 Dy 1 CR 19.8 +/- 0.7 6 Acfer 082 A 82 CV3 8.5 +/- 0.2 Acfer 086 A 86 CV3 15.4 +/- 0.3 Acfer 272 A 272 CV3 13.0 +/- 0.3 Acfer 182 A 182 CH 16.3 +/- 0.6 12.2 Acfer 207 CH 16.8 +/- 0.5 12.2 The 10Be and 26Al data of the two CO3 chondrites Acfer 202 and Acfer 243 indicate that the two meteorites are not paired. Based on the 10Be and 26Al data presented here, a pairing of the three meteorites Acfer 087, Acfer 186 and El Djouf 001 classified as CR seems very certain. This interpretation of the data is supported by results of mineralogical and petrographical investigations [8]. Remarkable about this pairing is the fact that the El Djouf location is more than 500 km apart from the Acfer location. The investigation on 53 H, L and LL meteorites from the Sahara by Wlotzka et al. (this issue) indicates that the terrestrial ages of 98 % of these meteorites are less than 40000 years. Assuming that this is true for C chondrites also, the terrestrial age of the CR chondrites is negligible compared to the halflives of 10Be and 26Al. So, the exposure age of 6 Ma implies that the 10Be and 26Al activities are in saturation and consequently represent production rates. In the case of 26Al, this is consistent with the activity range reported for other carbonaceous chondrites [11,12,13]. In the case of the CV3 condrites based on the 10Be and 26Al concentrations a pairing seems very improbable, this conclusion is supported by mineralogical and petrographical investigations [7]. The 10Be and 26Al concentrations of the two meteorites Acfer 182 and Acfer 207 classified as CH indicate a pairing; likewise do the results of mineralogical and petrographical investigations [9]. If the terrestrial age of these meteorites is negligible, in this case, too, regarding the high exposure age of 12.2 million years, the 10Be and 26Al activities appear to be saturated. Table 2 10Be and 26Al concentrations in mesosiderites and other classes from the Sahara*measured by means of AMS and gamma-gamma- coincidence-spectrometry Meteorite Sample Class 26Al [dpm/kg] gamma-gamma AMS Acfer 063 A 063 Meso 37.4 +/- 1.9 30.7 +/- 3.6 Silicate 67.0 +/- 3.4 Acfer 265 A 265 Meso 30.7 +/- 2.1 Ilafegh 002 Il 2 Meso 33.9 +/- 1.9 Silicate Acfer 217 A 217 R 56.5 +/- 3.4 64.1 +/- 6.4 Acfer 277 A 277 Ure <10 15.5 +/- 1.7 Ilafegh 009 Il 009 EL7/6 68.6 +/- 3.4 Tanezrouft 031 T 31 E 34.5 +/- 2.1 49.4 +/- 3.3 Acfer 193 LL4/6 55.7 +/- 2.8 Acfer 160 LL Brec. 45.4 +/- 2.3 Adzhi-Bogdo LL3/6 Brec. 56.4 +/- 2.8 Acfer 031 H5 50.0 +/- 2.5 Acfer 038 H5 52.9 +/- 4.2 Meteorite Sample Class 10Be [dpm/kg] T(rad) AMS [Ma] Acfer 063 A 063 Meso 7.0 +/- 0.2 7.3 +/- 0.2 Silicate 12.1 +/- 0.2 Acfer 265 A 265 Meso Ilafegh 002 Il 2 Meso 4.4 +/- 0.2 Silicate 8.0 +/- 0.3 Acfer 217 A 217 R 22.9 +/- 0.3 35 Acfer 277 A 277 Ure 2.6 +/- 0.8 2.6 +/- 0.5 Ilafegh 009 Il 009 EL7/6 18.2 +/- 0.3 18.7 +/- 0.4 17.3 +/- 0.7 Tanezrouft 031 T 31 E 13.0 +/- 0.5 Acfer 193 LL4/6 Acfer 160 LL Brec. Adzhi-Bogdo LL3/6 Brec. 18.1 +/- 0.5 Acfer 031 H5 Acfer 038 H5 * except for Adzhi Bogdo originating from Mongolia The 10Be data of the two mesosiderites Acfer 063 and Ilafegh 002 (table 2) and also the 10Be concentrations in the stone phases of both meteorites indicate that they are not paired. On the basis of the 26Al data a pairing of Acfer 063 and Acfer 265 seems probable, but the 10Be data are needed to make a more definite conclusion. In the case of Acfer 063 and Ilafegh 002, a magnetical separation of the iron phase from the stone phase was done to investigate the 10Be and 26Al concentrations in the stone phase separately. Taking the saturation activities of eucrites as references (21.8 +/- 0.7 and 93 +/- 14 dpm/kg, respectively [14]), the 10Be and 26Al concentrations in the stone phase of Acfer 063 appear to be undersaturated. Assuming these average saturation activities, the 10Be and 26Al concentrations imply an exposure age of 1-2 Ma. Acfer 217, classified as a member of the Rumuruti-type chondrite group (R), is another interesting sample that will be discussed here. Based on model calculations the following conclusions on Acfer 217 can be made: The concentrations of all cosmogenic nuclides can be consistently interpreted only if a preatmospheric radius between 15 and 65 cm, a single stage exposure history and a terrestrial age low compared to the half-life of 26Al are assumed [10]. The average 26Al and 10Be production rates in ureilites are 42.3 +/- 2.2 and 20.3 +/- 1.0, respectively [14]. Taking these production rates as references, the low 26Al concentration of 10 - 15 dpm/kg in Acfer 277 indicates an exposure age of 0.3 -0.5 Ma, the 10Be concentration of 2.6 dpm/kg consistently indicates an exposure age of 0.3 Ma. The 26Al concentrations of the H and LL chondrites as well as the 10Be concentration of Adzhi-Bogdo are within the normal range for these classes [14]. Acknowledgements. We thank the Deutsche Forschungsgemeinschaft for the support of this research. References [1] U. Herpers et al. (1967) IAEA Vienna, 199. [2] U. Herpers et al., Meteorite Research (1969) Reidel, Dordrecht, 387. [3] S. Vogt (1988) Thesis, Universitat zu Koln. [4] S. Vogt and U. Herpers (1988) Fresenius Z. Anal. Chem. 331, 186. [5] M. Suter et al. (1984) Nucl. Instr. and Meth. B5, 117 [6] H. A. Synal (1989) Thesis, ETH Zurich, ETH-Nr. 8987. [7] A. Bischoff, private communication. [8] A. Bischoff et al. (1993) GCA 57, 1587-1603 [9] A. Bischoff et al. (1993) GCA 57, 2631-2648. [10] A. Bischoff et al. (1994) Meteoritics 29, No. 2, 264-274. [11] U. Herpers and P. Englert (1983) Proc. LPSC 14th in JGR, 88, B312-B318. [12] K. Nishiizumi (1987) Nucl. Tracks Radiation Meas. 13, No.4, 209- 273. [13] U. Herpers et al. (1990) LPI Tech. Rept. 90-01, 46-48. [14] S. Vogt et al. (1990) Rev. Geophys., 28, 253-275. Welten K. C.* Alderliesten C. van der Borg K. Lindner L. Cosmogenic 10Be and 26Al in Lewis Cliff Meteorites 1. INTRODUCTION Terrestrial ages of Antarctic meteorites give information about the age of the stranding surface they are found on and may provide more insight into the concentration mechanisms involved. In this way, Antarctic meteorites contribute to a better understanding of the history of the Antarctic ice sheet [1], which is of special interest because most stranding surfaces are found in those regions where the ice sheet is very sensitive to climate changes [2]. One of the most 'productive' blue ice fields is the Lewis Cliff stranding area (84 degrees 15' S, 161 degrees 30' E), which is located between Law Glacier and Beardmore Glacier, two major outlet glaciers near the Queen Alexandra Range. So far, it has yielded more than 1800 meteorites, most of which were found on the Lewis Cliff Ice Tongue, a small blue ice field of 2.5 by 8 km. Earlier 26Al measurements on 30 meteorites [3] - as part of an ongoing gamma-ray survey - and 36Cl AMS measurements on 8 meteorites [4] suggest that the Lewis Cliff area is a relatively old stranding surface. Here we present 26Al gamma-ray results of 85 Lewis Cliff ordinary chondrites (OC's), obtained by measurements on bulk samples of 30 - 500 grams. These results will be compared with those of 557 OC's from the Allan Hills Ice Fields [5] in order to determine the relative age of the Lewis Cliff stranding area. The relationship between the 26Al activity and the location of find on the ice will be discussed, to shed more light on the local meteorite concentration mechanism. For 67 Lewis Cliff OC's the 26Al activities will be compared with available natural thermoluminescence (NTL) data [6] in order to test their previously suggested relationship (both 26Al and NTL decrease with terrestrial age) and to identify meteorites with unusual exposure histories. On the basis of our gamma-ray results, material was selected for subsequent AMS measurements of 10Be, 26Al and 36Cl. Here we present AMS results for 10Be and 26Al of 29 Lewis Cliff samples. The combination of 26Al with 10Be data allows a better estimate of the terrestrial ages of these meteorites, because corrections can be made for short exposure ages where necessary. Finally, we used our gamma-ray and AMS results to identify potential pairings. 2. EXPERIMENTAL Gamma-ray measurements. For bulk samples of 30 - 500 g gamma-ray spectra were measured for 5 - 15 days by high-purity Ge detectors of either 110 cm^3 or 90 cm^3. Efficiency calibration curves were made, based on 16 meteorite mock-ups of different size and shape. The mock-ups were made of a thin araldite shell, filled with a mixture of ointment and Cu- powder, thus matching both the specific gravity and electronic density of ordinary chondrites. Known amounts of KCl and 26Al were added to the mixture in order to calibrate the efficiency of the 1.46 MeV gamma-line of natural 40K and the 1.81 MeV gamma-line of cosmogenic 26Al. The calibration curve adds an uncertainty of 2 - 4 % to the statistical uncertainties, which are 5 - 10 % for 26Al and 2 - 5% for 40K. The natural 40K activities are used to deduce the K content of the meteorites. AMS sample preparation. Chondrite samples of about 1 g were crushed in an agate mortar. The metal phase was separated from the non-magnetic silicate fraction by hand with a magnet. The silicate fraction was further homogenized and used for 10Be and 26Al measurements, while the metal phase was kept for 36Cl analysis. We added 1.0 ml of beryllium and aluminium standard solutions (Merck, 1000 +/- 2 mg/l) to silicate subsamples of 40 - 60 mg. Dissolution of the samples and chemical preparation of BeO and Al2O3 targets was carried out as described previously [7]. AMS measurements. The 10Be and 26Al measurements were carried out at the Utrecht AMS facility [7]. Both isotopes were measured routinely using a fast beam-switching method, resulting in precisions of 1 - 2% for the 10Be/9Be ratios and 2 - 4% for the 26Al/27Al ratios. In addition, the precision of the resulting 26Al content is affected by the uncertainty in the indigenous 27Al content of a meteorite sample. From bulk Al contents [8] and a correction for average metal contents of 9 % for L- chondrites and 17% for H-chondrites, we estimated the average Al-content of the silicate fraction of H- and L-chondrites at 13.5 +/- 1.3 mg/g. The addition of extra Al carrier reduced the uncertainty in the total amount of 27Al to 4 %. 3. RESULTS AND DISCUSSION 3.1 26Al gamma-ray survey General distribution of 26Al contents. Table 1 lists the measured 26Al gamma-ray activities and K contents of 85 Lewis Cliff OC's. Figure 1 shows the distribution of their 26Al activities, compared with that of 557 specimen from the Allan Hills region [5]. All H-chondrite activities were normalized to L-chondrite composition by multiplying with a factor of 1.07. It should be noted that neither histogram has been corrected for pairings. The 26Al activities range from 23 to 75 dpm/kg, except for two meteorites with values of 7 and 96 dpm/kg, respectively, which are unusual in comparison to the average saturation value of 60 dpm/kg. These and other anomalous cases will be discussed separately. The Lewis Cliff distribution of 26Al activities shows a broad peak between 35 and 55 dpm/kg, well below the saturation value of 60 dpm/kg. Although some of these low values might be due to short exposure ages - as will be shown later - or irradiation in space as small objects, it seems plausible that many of these meteorites have terrestrial ages of 100 ka or more. Figure 1 also illustrates that the Lewis Cliff OC's peak at lower 26Al activities than the Allan Hills OC's, suggesting that the Lewis Cliff stranding area harbors on the average older meteorites and thus has been in operation for a longer period than the Allan Hills area, the oldest stranding surface thus far known. Interestingly, most of the meteorites from the Lewis Cliff Ice Tongue were found on the western side, on relatively old ice, which was formed under the very cold climatic conditions of a glacial periods, as indicated by its very low delta^(18)O values (-53 to -58 per mil) [9]. However, for the Lewis Cliff meteorites investigated thus far, we found no obvious relationship between their 26Al activity and the location on the ice field [10]. This hints at a complex glaciology of the Lewis Cliff stranding area, as was suggested previously on the basis of field observations [1]. NTL vs. 26Al contents. Thermoluminescence studies [6] were carried out on 67 of the 85 Lewis Cliff samples that we measured in our gamma-ray survey. Figure 2, which shows the combined 26Al (L-normalized) and NTL results for these samples, reveals two main groups: one with high NTL values (> 5 krad) and one with low NTL values (< 2 krad). The high-NTL group exhibits two outliers, one with a 26Al activity of 7 dpm/kg and one with 75 dpm/kg. The low-NTL group, however, contains a relatively high number of 3 (out of 9) meteorites with unusual, 'over-saturated' 26Al values, as will be explained later. Except for the outliers, the high-NTL group shows 26Al values between 20 and 60 dpm/kg. Contrary to earlier reports [6], there is no obvious correlation between NTL values and 26Al activities. This indicates that for Lewis Cliff meteorites the NTL value is not even roughly related to the terrestrial age. More recently, it was indeed argued that the NTL level of an Antarctic meteorite is largely determined by its 'surface exposure age' rather than its total terrestrial age [12]. This still does not explain the negative - rather than positive - correlation we observe between NTL and 26Al in Lewis Cliff meteorites. Unusual exposure histories. The lowest 26Al activity of 7 +/- 2 dpm/kg found in LEW 86360 (L4) hints at either a very high terrestrial age (>= 2 Ma), an extremely short exposure age (<< 1 Ma) or a complex exposure history. The latter explanation seems most likely, as was also concluded for the Antarctic meteorite ALH 76008 with a similarly low 26Al activity of 11 +/- 1 dpm/kg [11]. 10Be and noble gas measurements are currently in progress to shed more light on this problem. The low-NTL group of figure 2 contains three L6 chondrites (87143, 88190 and 87169) with 'over- saturated' 26Al activities, ranging from 69 to 96 dpm/kg. These high 26Al values can be explained by a contribution of solar cosmic-ray (SCR) produced 26Al, indicating a low-perihelion (< 0.85 AU) orbit. During low-perihelion passage meteorites are also exposed to solar heating, which explains the low NTL values. High SCR fluxes result in increased 26Al contents in the outer few cm of a meteoroid and will therefore only be observed in those low-NTL meteorites which also suffered low atmospheric ablation. The high 26Al value of LEW 88013 does not coincide with a low NTL value and can therefore not be unambiguously attributed to SCR-produced 26Al. A large pre-atmospheric size (r = 40-50 cm) would also explain this high 26Al value [13]. 3.2 AMS results of 10Be and 26Al The 10Be and 26Al results of 29 Lewis Cliff meteorites and the Utrecht L6 chondrite as a modern reference (fall, 1843) are listed in table 2. For eight meteorites both a bulk sample and a silicate fraction were measured to check the consistency of our normalization procedure. This procedure converts values obtained on silicate samples into bulk values by taking into account the metal/silicate ratios and the 10Be and 26Al production rates for metal and silicate fractions. The comparison shows that the procedure is adequate within the error limits. Twenty-six of the meteorites listed in table 2 were also measured in our gamma-ray survey (table 1). Except for LEW 86534, which shows a gamma-ray 26Al activity of 37 +/- 4 dpm/kg versus an average AMS value of 53 +/- 2 dpm/kg, the results fallt on a line with a slope of 1.03 +/- 0.02. Terrestrial ages of Antarctic meteorites can be deduced from their 26Al acitivity (A26) by assuming a L- normalized saturation value of 60 dpm/kg and a half-life of 705 ka. However, for a meteorite that never reached the saturation level, the resulting value is an 'apparent' terrestrial age which must be corrected for its short exposure age. Due to the longer half-life of 10Be (t 1/2 = 1.51 Ma) a correction can be made on the basis of the 10Be activity. In figure 3 we plotted A26/A10, the measured 26Al/10Be activity ratio, versus the 'apparent' terrestrial age of each meteorite. Whereas the individual 10Be and 26Al production rates vary up to 30 % due to shielding effects, their ratio appears to show variations of less than 10 % for objects within a wide range of pre-atmospheric radii [13]. With an L-normalized saturation value of 20 dpm/kg for 10Be, the A26/A10 ratio at saturation is 3.0 +/- 0.3, as is indeed found for the Utrecht chondrite (table 2). Higher ratios are the result of short exposure ages (< 5 Ma), lower ratios are due to radioactive decay of 26Al and 10Be during residence on Earth. This dependence of the A26/A10 ratio on exposure age and terrestrial age is respresented in figure 3 by a two-dimensional grid. The terrestrial age coordinate of this grid represents the change in time of the A26/A10 ratio, which decreases by a factor of two each 1.32 Ma. By following for each data point the directions of the grid lines one arrives at the corresponding values for the exposure age and the terrestrial age, respectively. The corrected terrestrial ages thus obtained for the 29 samples plotted in figure 3 range up to 500 ka. The terrestrial age distribution (figure 4) corrected for two pairing groups (no. 13 and 14) shows a peak at low values (< 100 ka) as well as a small peak between 150 and 250 ka. The general shape of this distribution is similar to that of Nishiizumi's distribution of terrestrial ages based on AMS 36Cl measurements of 28 Lewis Cliff meteorites [14]. However, in the latter distribution the peak at low ages is more pronounced (64 % of the cases vs. 29% in our data). We should point out that this difference might be explained by shielding effects, because the terrestrial ages of meteorites with small pre- atmospheric sizes (r < 15 cm) are still overestimated in our approach. Additional 36Cl measurements (in progress) will provide more definite answers. Two meteorites plotted in figure 3 show A26/A10 ratios well above the grid. Because one of these (LEW 87169) most likely contains SCR-produced 26Al, this might also be the case for the other meteorite (LEW 86016), although its 26Al activity of 64 dpm/kg does not significantly hint at 'over-saturation'. More cosmogenic nuclide data are required to shed light on its (unusual ?) exposure history. Figure 3 reveals a cluster of five meteorites with short exposure ages of about 1 Ma. The upper four, L6 chondrites from Meteorite Moraine, are probably - in view of their similarly low 10Be and 26Al values (table 2) - fragments of a single fall. The fifth meteorite, an H5 chondrite, reveals a terrestrial age of 330 +/- 60 ka, which is in agreement with the independently determined 36Cl age of 300 +/- 60 ka [15]. 3.3 Pairing. Several potential pairings have been suggested by others on the basis of location, chemical-petrological classification and NTL results [5,6]. Cosmogenic nuclide data provide additional and important clues in identifying or excluding pairings. Also the K contents obtained in our gamma-ray survey serve as an additional pairing criterion. Of the 85 meteorites measured, 54 specimen could be assigned to 16 different pairing groups, each with 2-6 members, as indicated in table 1. Of these 16 groups, only one pairing (88019 and 88020) was already suggested on the basis of their classification and proximity on the ice field [5]. Some of the pairings suggested on the basis of TL and other data [6] were confirmed, whereas others could be rejected. Due to analytical uncertainties and shielding effects it is usually difficult to identify pairs with absolute certainty. Therefore we used the confidence levels a (probably paired), b (tentatively paired), c (possibly paired) and x(unpaired) introduced by Scott [16]. We assigned two groups with confidence level a), five with level b) and nine with level c). Since for the remaining 31 meteorites pairing can be excluded, the 85 specimen represent at least 47 falls. The upper limit for individual falls amounts to 70 if only the a- and b-pairings are taken into account. Since some of the assigned c-pairings may later, when more data will become available, prove to represent individual falls, we estimate the total number of individual falls between 50 and 60. This brings the average number of fragments per fall for the Lewis Cliff area at 1.5. However, for two reasons this value must be considered as a lower limit: (i) our sample selection is biased to specimen larger than 50 grams, whereas it is known that the number of fragments increases towards smaller sizes and (ii) we only considered pairings between meteorites found fairly close to each other, that is on the same field, either Lower Ice Tongue, Upper Ice Tongue, Meteorite Moraine, South Lewis Cliff or Upper Walcott Neve. This implies that the average number of fragments per fall for the Lewis Cliff stranding area will be close to or within the range of 2-6 fragments per fall, as estimated by Scott on the basis of 300 Allan Hills meteorites [16,17]. References: [1] Cassidy W. et al. (1992) Meteoritics 27, 490-525; [2] Delisle G. (1993) J. Glaciol. 39, 397-408 [3] Welten K.C. et al. (1992) Meteoritics 27, 306-307; [4] Nishiizumi K. et al. (1991) Meteoritics 26, 380; [5] Grossman J.N. (1994) Meteoritics 29, 100-143; [6] Benoit P.H. et al. (1992) J. Geophys. Res. 97, no. B4, 4629-4647; [7] Welten K.C. et al. (1994) Nucl. Instr. and Meth. in Phys. Res. B92, 500-504; [8] Kallemeyn G.W. et al. (1989) Geochim. Cosmochim. Acta 53, 2747-2767; [9] Faure G., et al. (1993) Antarctic J. of the U.S. 28 (5), 69-70; [10] Schutt J. and Fessler B. (1991) Antarctic Meteorite Location and Mapping Project, LPI, Houston, Texas [11] Nishiizumi K. et al. (1979) Earth Planet. Sci. Lett. 45, 285- 292; [12] Benoit P.H. and Sears D.W.G. (1994) in 'Workshop on Meteorites from Cold and Hot Deserts'; [13] Vogt S. (1990) LPI Tech. Rept. 90-05, 112-118; [14] Nishiizumi K. (1994) in 'Workshop on Meteorites from Cold and Hot Deserts'; [15] Nishiizumi K., priv. communication; [16] Scott E.R.D. (1984) Mem. NIPR Spec. Issue 35, 102-125; [17] Scott E.R.D. (1989) Smithsonian Contrib. Earth Sci. 28, 103-111 Reedy R. C.* Masarik J. Production Profiles of Nuclides by Galactic-Cosmic-Ray Particles in Small Meteoroids Many of the meteorites found in cold and hot deserts are small, and many were small bodies in space. Production of cosmic-ray-produced (cosmogenic) nuclides in small meteoroids is expected to be different than that in the larger meteoroids typically studied [1], with lower levels of nuclide production by galactic-cosmic-ray (GCR) particles [e.g., 2] and possibly significant production by solar-cosmic-ray (SCR) protons [3]. Motivated by the cosmogenic-nuclide measurements for the very small Salem meteorite [4,5] and for cosmic spherules [e.g., 6], which show high levels of SCR production, we have reported earlier nuclide production rates by SCR protons in small objects in space [e.g., 3,7]. The GCR production rates reported in [2] for small meteoroids have not been tested and were expected to be poor for meteoroids with radii <40 g/cm^2 because of the very simple nature of that semi-empirical model (only one free parameter) and because the mix of neutrons and protons is different (relatively more protons) than that in the model, which was based on larger objects. Thus we have calculated production rates for nuclides made by GCR particles in small objects with a physical model that is much better suited for unusual targets. The production rates for GCR nuclides were calculated using particle fluxes from the Los Alamos Monte Carlo LAHET Code System (LCS) and measured or evaluated cross sections [8]. LCS has yielded calculated production rates that almost always are in very good agreement with cosmogenic-nuclide measurements in meteorites with radii greater than about 15 cm [8--10]. The fluxes of protons and neutrons in spherical meteoroids of radii from 1 to 45 cm and with an L-chondrite composition were calculated as a function of pre-atmospheric depth. Production rates for most stony objects are similar to those for L-chondrites [8]. The neutron and protons fluxes calculated by LCS were normalized to an effective incident omnidirectional GCR proton flux of 4.8 protons/(cm^2 s) [9]. This flux is ~60% higher than that for the primary protons in the GCR as it includes production by alpha particles and heavier GCR nuclei, both directly and by secondary nucleons contributed by the break-up or reactions of these heavier GCR nuclei. We have found that the GCR flux averaged over the orbits of meteoroids are ~5% greater than that at the Moon [11]. For each layer, these fluxes were then multiplied by the neutron and proton cross sections for major target elements and integrated over energy to get the production rates for eight nuclides: ^10Be, ^14C, ^21Ne, ^22Ne, ^26Al, ^36Cl, ^38Ar, and ^53Mn. The calculated GCR production rates for the smallest meteoroids are ~70% greater than those calculated using only primary GCR protons by [12], mainly because of the ~60% factor for heavier GCR nuclei, which do not break-up or react in very small meteoroids. Our calculated GCR production rates for radii up to ~10 cm also could be high. The only test of GCR production rates in very small meteoroids is the ^10Be activities in Salem, which are ~=17 dpm/kg [5], in good agreement with our ^10Be production rates calculated for a 3-cm-radius meteoroid. Thus the most serious errors in our calculated production rates are for radii less than ~3 cm. The plots in Fig. 1 show the results of our production-rate calculations. All of the cosmogenic nuclides have GCR production profiles that increase from the pre-atmospheric surface to a maximum near the center of the meteoroid. The results are basically similar to those in [2]. The amount of this increase varies considerably, with products made by higher-energy particles and by protons, such as ^10Be, having less increase in production rate with depth. In all cases, the GCR-calculated production rates in meteoroids with radii less than ~15 cm are much less than those for larger meteoroids. As most cosmogenic- nuclide production rates have been inferred from measured concentrations in such larger meteoroids, their use for small meteoroids results in over-estimated production rates. For example, our calculated ^21Ne production rate for Salem as a 3-cm sphere and the measured ^21Ne concentration [13] yields an exposure age for Salem of about 15 Ma, more than the 9.5-Ma age reported in [5] using systematics based on larger objects, although our calculated ^22Ne/^21Ne ratio (~1.35) is higher than the observed ratio of 1.23. However, this is an upper limit to the age as SCR contributions were not included. Also, possible complications, such as erosion in space [14], have also not been considered. Except probably for ^10Be, which has quite low SCR-production rates [15], it should be remembered that nuclide production by SCR particles should be added to the GCR contribution in interpreting measured concentrations of cosmogenic nuclides. These SCR contributions can be significant, especially near the pre-atmospheric surface and in the smaller meteoroids [3,7]. The amount of the SCR contribution is not only sensitive to the pre-atmospheric radius of the meteorite and a sample's pre-atmospheric depth [3,7], but also to the orbit of the meteoroid prior to hitting the Earth. The time-averaged flux of SCR particles decreases rapidly with increasing distance from the Sun, as R^-2 because of the distance from the Sun and possibly with poorly-known additional factors of up to R^-0.5 or more due to interactions of SCR particles with the interplanetary medium [16]. References: [1] Reedy R.C. (1990) Meteoritics 25, 400. [2] Reedy R.C. (1985) Proc. 15th Lunar Planet. Sci. Conf., C722. [3] Reedy R.C. (1987) Lunar Planet. Sci. XVIII, 822. [4] Evans J.C. et al. (1987) Lunar Planet. Sci. XVIII, 271. [5] Nishiizumi K. et al. (1990) Meteoritics 25, 392. [6] Nishiizumi K. et al. (1990) Earth Planet. Sci. Lett. 104, 315. [7] Reedy R.C. (1990) Lunar Planet. Sci. XXI, 1001. [8] Masarik J. and Reedy R. C. (1994) Geochim. Cosmochim. Acta, in press. [9] Reedy R.C. et al. (1993) Lunar Planet. Sci. XXIV, 1195. [10] Englert P.A.J. et al. (1995) Geochim. Cosmochim. Acta, in press. [11] Reedy R.C. and Masarik J. (1994) Lunar Planet. Sci. XXV, 1119. [12] Reedy R.C. (1987) Proc. 17th Lunar Planet. Sci. Conf., E697. [13] Wieler R. (1992) Priv. Comm. [14] Reedy R.C. (1992) Meteoritics 27, 280. [15] Nishiizumi K. et al. (1988) Proc. 18th Lunar Planet. Sci. Conf., 79. [16] Hamilton D.C. (1988), In Interplanetary Particle Environment, Jet Propulsion Laboratory Publ. 88--28, p. 86. * Work supported by NASA and done under the auspices of the U.S. Dept. of Energy. Figure 1. Calculated GCR production rates (or ratios) as a function of depth in meteoroids of various pre-atmospheric radii. Thursday, July 21, 1994 TERRESTRIAL AGES 1:30 - 3:00 p.m. Chair(s): M. E. Lipschutz Jull A. J. T.* Bevan A. W. R. Cielaszyk E. Donahue D. J. 14C Terrestrial Ages and Weathering of Meteorites from the Nullarbor Region, Australia The time a meteorite resides on the Earth's surface, its terrestrial age [1-3] is important in determining the history of the meteorite. 14C was originally used for large samples, but for the last ten years, smaller samples of meteorites of 0.1 to 0.5g have been dated using accelerator mass spectrometry [2-5]. The precision of terrestrial age estimates is limited by the accuracy to which the saturated activity of 14C in the meteorite is known, and the production of 14C can vary with the depth and size of the object. 14C as a function of accurate depth is known for the L5 chondrite, Knyahinya [6]. We have used Knyahinya, Bruderheim and some other chondrites to establish a saturated activity reference level of about 51 dpm/kg for L-chondrites [3,6]. The storage time before a meteorite weathers away is much less for warm, arid regions than in some areas of Antarctica, and 14C (t1/2 5,730 years) is the ideal radioisotope to use for estimates of terrestrial age. As weathering gradually destroys meteorites in a given population, the resulting distribution for similar types of meteorites should be an approximately exponential decrease of meteorites with increasing age. Boeckl [7] determined a "weathering half life" of some 3,500 years for chondrites from the southwestern USA, but a later reinvestigation of this study [3] determined that the 14C age distribution of the meteorites was longer. We have studied 14C ages of meteorites from Roosevelt County, New Mexico [8], the western Libyan desert [5] and Algeria [9]. In these areas meteorites of ages as old as >40,000 yr are observed, and the mean survival time at these locations is well over 10,000 yr. For 13 meteorites from the Hammadah al Hamra, Libya, Jull et al. [5] found a break in the age distribution which might be related to the timing of climatic changes in the collection area. The Roosevelt County collection also shows departures from exponential behavior, possibly due to storage of the meteorites in Quaternary cover sands in the area [8]. More than 3.8 million km^2 of Australia is arid or semi-arid land that provides conditions for the prolonged preservation of meteorites [10,11]. Two areas in particular have been recognized as areas containing high accumulations of meteorites, and Bevan and co-workers [10-13] have documented large numbers of finds in the Nullarbor Region of Western Australia. In this paper, we have studied the 14C age distribution of over 20 meteorites from the Nullarbor Region, Western Australia. Figure 1 presents the 14C age distribution of Nullarbor samples compared to some other arid locations where a substantial number of 14C ages have been obtained. The Western Australian meteorites show a simple exponential dependence of terrestrial age versus time, and no meteorites of greater than about 30Kyr age. This is in contrast to the results from the southwestern USA [3], Roosevelt County [8], Algeria [9] and Antarctica [14]. One might expect that meteorites would be more well- preserved in a very arid, hot climate. However, the lack of very old stony meteorites in the Nullarbor compared to other locations may be solely a statistical problem. Weathering of meteorites in arid regions is expected to be dominated by the availability of moisture, though could be accelerated by the presence of chlorine [15]. Episodic heavy rainfall events probably provide the main source of water, which causes weathering. We have also studied the carbonates in the weathering products of some of these meteorites. These results show that there are some variations in delta ^13C, and there is a weak correlation of delta ^13C and carbonate content with terrestrial age. The expected exponential drop- off of number of meteorites of a given terrestrial age with time indicates the collection area has been substantially undisturbed during at least the last 30,000 years. This is certainly consistent with the arrested karst geomorphology of the Nullarbor. The surface of the Nullarbor is considered to be eroding slowly [16], and it has probably been generally stable for considerably longer than 30,000 years. This is not seen in the US meteorite collections from Roosevelt County or Northwest Texas. The less arid and colder high plains of Texas and New Mexico may result in storage of meteorites by burial for longer periods of time. We observe a deficit of "young" meteorites for these areas. In the sample of meteorites analyzed so far, the Nullarbor data shows no evidence of selection of meteorites of a particular terrestrial age. The Nullarbor may prove to be an important area to provide data on the flux of meteorites with time. Although cyclonic winds have moved small fragments, more than 800 fragments of the Mulga (North) shower, with a terrestrial age of 2.7+-1.3 Ka, were found in a well- defined ellipse [17]. Concentrations of meteorites, such as in the Nullarbor Region, also provide datable materials of a variety of terrestrial exposure times during the accumulation period. The degree of weathering of meteorites of different ages may allow changes in weathering rates with time and perhaps even some climatic effects to be estimated. References: [1] Nishiizumi K. et al., EPSL, 93, 299. [2] Jull A. J. T. et al. (1989), GCA, 53, 2095. [3] Jull A. J. T. et al. (1993), Meteoritics, 28, 189. [4] Beukens R. P. et al. (1988), Proc. NIPR Symp. Ant. Met., 1, 224. [5] Jull A. J. T. et al. (1990), GCA, 54, 2895. [6] Jull, A. J. T. et al. (1994), Meteoritics, 29, 649. [7] Boeckl, R. P. (1972) Nature, 236, 25 [8] Jull A. J. T. et al. (1991) LPSC-22, 665. [9] Wlotzka F. et al., (1994), Workshop on Meteorites from Cold and Hot Deserts, N"rdlingen, LPI Tech. Rep. [10] Bevan, A. W. R. (1992), Records of the Australian Museum, Supp. 15, 1. [11] Bevan, A. W. R. et al. (1992) Meteoritics, 27, 202. [13] Bevan, A. W. R. and Binns, R. A. (1989), Meteoritics, 24, 127 and 24, 134. [14] Jull A. J. T. et al. (1993) Meteoritics, 28, 376 [15] Buchwald V. F. and Clarke R. S. Jnr (1989), Amer. Min., 74, 656. [16] Gillieson D.S. and Spate A. (1992) In Gillieson, D. S. (ed) Special publ 4, Dept Geography and Oceanography Univ. College, Austr. Defence Acad., Canberra, 65-99. [17] Cleverly, W. H. (1972) J. Roy. Soc. West. Austr. 55, 115. Figure 1: Distribution of 14C terrestrial ages from Western Australia and other arid regions. Table 1: 14C terrestrial ages of meteorites from the Nullarbor Plain, Australia Sample 14C dpm/kg Terrestrialage(yr) Billygoat Donga,L6 20.5 +- 0.3 7,550 +- 1300 20.0 +- 0.2 7,700 +- 1300 Boorabie 001, H4-5 41.7 +- 0.3 900 +- 1300 Burnabbie, H5 2.85 +- 0.2 23,100 +- 1400 Carlisle Lakes 002, H4-5 32.9 +- 0.3 2,800 +- 1300 Cocklebiddy, H5 36.9 +- 0.1 1,900 +- 1300 Deakin 001, anomal. 1.9 +- 0.1 27,100 +- 1400^1 acid residue 2.4 +- 0.4 25,300 +- 1900^2 Forrest 007, H4 30.9 +- 0.3 3,400 +- 1300 Forrest 009, L6 25.0 +- 0.2 5,900 +- 1300 Forrest 010, L4-5 5.84 +- 0.09 17,900 +- 1300^3 acid residue 1.4 +- 0.3 29,500 +- 2200^(2,3) Kybo 001, LL5 42.5 +- 0.3 2,200 +- 1300 Mulga (north), #417, H6 33.4 +- 0.2 2,720 +- 1300 Mulga (north), #585, H6 34.4 +- 1.0 2,500 +- 1300 Mulga (south), H4 4.1 +- 0.2 20,000 +- 1300 Mundrabilla 002, H5 2.46 +- 0.09 24,300 +- 1330 Mundrabilla 005, H5 52.2 +- 0.6 recent fall North Forrest, H5 10.2 +- 0.9 11,980 +- 1300 acid residue 11.9 +- 0.3 11,200 +- 1300 North West Forrest, E6 31.1 +- 0.3 1,720 +- 1300 Nyanga Lake 001, H3 21.1 +- 0.2 6,500 +- 1300 Reid 006, H5 33.8 +- 0.5 2,600 +- 1300 32.9 +- 0.2 2,840 +- 1300 Reid 007, L6 23.8 +- 0.2 6,300 +- 1300 Reid 010, H6 39.3 +- 0.3 1,400 +- 1300 Reid 011, H3-6 39.5 +- 0.3 1,300 +- 1300 1 L-chondrite composition assumed in the calculated of terrestrial age. 2 The sample was treated with phosphoric acid to remove any weathering carbonates. 3 The first measurement of Forrest 010 released a large amount of carbon. The acid-treated sample of Forrest 010 indicates that 14C from weathering carbonates was removed during pretreatment. Figure 1, showing the distribution of 14C terrestrial ages from Western Australia and other arid regions, appears here in the hard copy. Nishiizumi K.* Terrestrial Ages of Meteorites from Cold and Cold Regions We are continuing our ongoing study of cosmogenic nuclides in Antarctic meteorites. The major objective is the determination of the terrestrial ages of meteorites based on 36Cl concentrations. The distribution of meteorite terrestrial ages is one of the most vital pieces of information for studies of ice movement and meteorite accumulation mechanisms in the blue ice areas. In addition to work on terrestrial ages, we are also studying the histories of Antarctic meteorites and of the cosmic rays. We have measured 36Cl in over 150 Antarctic meteorites since our previous publication [1]. Although much of the new data is still preliminary, some interesting points are already evident. We present here findings and new observations based on both our new results and on previous studies. Since a large number of meteorites have been recovered from many different icefields in Antarctica, we continue to survey the trends of terrestrial ages for different icefields. We have also measured detailed terrestrial ages vs. sample locations for Allan Hills, Elephant Moraine, and Lewis Cliff Icefields where meteorites have been found with very long ages. Fig. 1a and 1b show the updated histogram of terrestrial ages of meteorites from the Allan Hills Main Icefield, Allan Hills Far Western and Middle Western Icefield, Yamato Mountains, Elephant Moraine, Lewis Cliff, and other regions in Antarctica. The figure includes 14C and 81Kr terrestrial ages obtained by other groups. Pairs of meteorites are shown as one object plotted at the average. The width of the bars represents 70,000 years which is a typical uncertainty for 36Cl ages. The total amount of data has more than doubled since our previous publication [1]. The Allan Hills Icefields and the Allan Hills meteorites are the most intensively and widely studied to date. We have therefore concentrated our studies in this area. Fig. 1a clearly indicates that meteorites found at the Allan Hills Icefields are much older than any other meteorites. Fig. 1a also shows a comparison of the terrestrial ages at both the Allan Hills Main Icefield and with the subsidiary surrounding icefields. Very old meteorites are only found on the Allan Hills Main Icefield. The terrestrial ages cover a wide range and are as old as 1 million years. Delisle and Sievers [2] have performed detailed field studies of ice topography and bedrock topography in the vicinities of the Allan Hills Main Icefield and the Near Western Icefield. Many of the old terrestrial age meteorites were found over ice with shallow depths (over mesa like bedrock topography). Four Allan Hills meteorites (ALH 84243, 85037, 85048, and 85123) were collected on soil or on bedrock. Three of these have terrestrial ages less than 100,000 years but one (ALH 85048) has a 920,000 year terrestrial age. We do not yet understand the relationship between the terrestrial ages and the histories of the outcrops on which the meteorites were found. It is very important to study the exposure age of the bedrock using in-situ produced cosmogenic nuclides [3]. Many of the Lewis Cliff meteorites are as old as the Allan Hills (Main Icefield) meteorites. Eight out of 28 Lewis Cliff meteorites have terrestrial ages greater than 200,000 years. So far, no clear correlation has been found between the terrestrial ages and the locations of the Lewis Cliff meteorites. Old and young meteorites were fond on both the Lower and Upper Ice Tongue (see Fig. 1b). The terrestrial age determination of Lewis Cliff meteorites is related to two other projects. We have measured in situ produced 10Be and 26Al in a series of rocks between the Lewis Cliff Icetongue and Law Glacier in collaboration with Dr. G. Faure to investigate the progressive thinning of the East Antarctic ice sheet. The other project is measurement of 10Be, 26Al, and 36Cl in two "horizontal ice cores" which were collected by ANSMET during meteorite search at Lewis Cliff Icetongue. We will study the ice cores in collaboration with glaciologists to understand the ice flow at the Lewis Cliff Icefield and to investigate meteorite accumulation mechanisms. Although there is no clear correlation between the terrestrial age and weathering category [1], we compare the abundance of metal in chondrites and terrestrial ages of these objects. The abundance of metal yields direct information of chemical weathering in Antarctica. The Fig. 2 shows the comparison of clean metal abundance in Antarctic H chondrites at different ice fields and Non-Antarctic H. Abundance of clean metal in Antarctic meteorites is significantly lower than non-Antarctic. Although the precision of our 36Cl measurement is now 1-3 %, the limitation of 36Cl terrestrial ages is the longer half-life, 300,000 year, and the range of saturation value. We measured the 36Cl saturation values in 26 chondrites with known terrestrial ages. Previously we used the saturation value 22.8 +/- 3.1 dpm [1]. In this study, We found a new saturation value 22.1 +/- 1.4 dpm (1 sigma). This is in a good agreement with the previous value but smaller error. Acknowledgments: I wish to thank A. J. T. Jull for providing unpublished 14C results. The 36Cl measurements were performed at the University of Rochester and Lawrence Livermore National Laboratory. This work was supported by NASA and NSF grants. References [1] Nishiizumi K. et al. (1989) Earth Planet. Sci. Lett. 93, 299-313. [2] Delisle G. and Sievers J. (1991) J. Geophys. Res. E96, 15577-15587. [3] Nishiizumi K. et al. (1991) Earth Planet. Sci. Lett. 104, 440-454. Fig. 1a. A histogram of terrestrial ages of Allan Hills Main Icefield, Allan Hills Far and Middle Western Icefield, and Yamato meteorites. The width of the bars represents 70,000 years. Fig. 1b. A histogram of terrestrial ages of Elephant Moraine, Lewis Cliff, and other Antarctic meteorites. The width of the bars represents 70,000 years. Figure 2. Comparison of clean metal abundance in Antarctic H chondrites at different icefields and Non-Antarctic H. Wieler R.* Caffee M. W. Nishiizumi K. Exposure and Terrestrial Ages of H Chondrites from Frontier Mountain Meteorite populations from geographically distinct locations in Antarctica have different terrestrial age distributions. For example, meteorites found in the Allan Hills main icefield hav ages of up to a million years, whereas Yamato specimens are rarely more than 200,000 years old [1-3].Prior to this work, only few terrestrial age data have been reported for meteorites from the Frontier Mountain location (FRO) in North Victoria Land. Delisle et al. [4] give terrestrial ages for three FRO meteorites found during the 1984/85 field season, based on 26Al concentrations: FRO8401, (3.8 +/- 1.3)x10^5 a; FRO8403, (7.1 +/- 1.3)x10^5 a; FRO8421, (3.0 +/- 1.3)x10^5 a. These researchers observed from their limited data set that FRO meteorites displayed a similer range in residence times on Earth as other samples collected in Victoria Land. However, the one reported terrestrial age of a Frontier Mountain specimen based on 36Cl disagrees with its 26Al derived value: Nishiizumi et al. [1] state for FRO8403 an age of (1.2 +/- 1.0)x10^5 a, nearly seven times lower than the value given in [4]. Due to its half-life of 301'000 a 36Cl is better suited than 26Al to determine terrestrial ages in the 10^5 a range. We therefore initiated a study of terrestrial ages of FRO meteorites collected in the 1990/91 season by analyzing 36Cl in their metal fraction. We chose H chondrites of types 5 and 6 since this class yields the maximum amount of metal. The Frontier Mountain area and the find locations of the meteorites are described in [5]. Specimens labelled "Ice" in Table 1 were recovered on blue ice NE of Frontier Mountain, specimens labelled "Mor." are from "Meteorite Moraine". Regional ice flow and the probable meteorite concentration mechanism are discussed in [4]. We present He, Ne, Ar as well as 36Cl data of 12 Frontier Mountain meteorites. Judged by these data, the twelve samples seem to represent at least eight different falls. Pairing of the two solar gas rich samples FRO90002 and FRO90043 cannot be excluded and a common fall of FRO90001, FRO90050, FRO90073, and FRO90152 is suggested by their low 4He concentrations and low (3He/21Ne)cos ratios together with their similar 21Necos and (22Ne/21Ne)cos values. The parent meteorite of these four samples probably lost cosmogenic and radiogenic He. For each of the other specimens, find location, 36Cl activity, and/or noble gas signature suggests an individual fall. 21Ne exposure ages are calculated according to Graf et al. [6] and Eugster [7], in both cases using the (22Ne/21Ne)cos ratio as shielding parameter. No shielding information is available for the two solar gas bearing samples, for which we therefore assumed "average" shielding, i.e., (22Ne/21Ne)cos = 1.11, and report the exposure ages in parentheses. Ages derived by the two methods agree well for samples with (22Ne/21Ne)cos >= 1.09. This means that both methods assume a similar dependence between the 21Ne production rate (P21) and (22Ne/21Ne)cos for relatively low shielding. To derive P21 with the Graf et al. method, we assumed an upper limit of 30 cm for the preatmospheric radius "R" of all meteorites with (22Ne/21Ne)cos <=1.09. If we assumed larger preatmospheric radii, the upper bound of the exposure age interval derived for each sample would be correspondingly higher. The Eugster method does not explicitly make a similar assumption about R. However, since the variation of P21 with (22Ne/21Ne)cos used by Eugster [cf. ref. 8] has been derived with data from more than 100 meteorites which in their majority can be assumed to have had a preatmospheric radius of less than 30 cm, the Eugster formula is valid also only for meteorites within this limited size range. Samples with (22Ne/21Ne)cos <= 1.08 had a preatmospheric radius exceeding 30 cm, according to the Graf et al. model. We present only minimum ages for these four samples, which may all belong to the same fall. These minimum ages are up to 2.2 times higher than the respective values derived by the Eugster method. One reason for this discrepancy is that the P21 versus (22Ne/21Ne)cos correlation used by Eugster fails for (22Ne/21Ne)cos <= 1.08. Instead of becoming ever larger with increasing shielding, P21 will actually decrease in very big meteorites. However, it is not clear either how well the Graf et al. model extrapolates towards heavy shielding. It seems possible that the maximum 21Ne production rates around (22Ne/21Ne)cos = 1.05-1.06 are underestimated by the model and that the minimum exposure ages thus become too high. In summary, 21Ne exposure ages derived from heavily shielded samples are still quite uncertain. We therefore exclude the FRO90001 clan from the following discussion. Of the remaining eight samples in Table 1, only one has an exposure age clearly outside the 7 Ma peak so conspicious for H-chondrites [9, 10], and the exception (FRO90048) coincides with the other well established H-chondrite peak at about 33 Ma [10]. It is ambigous how many of the remaining samples really belong to the 7 Ma peak. Marti and Graf [10] note another peak at ~4 Ma for H5 and H6 chondrites, and indeed, two or three of our samples may belong to this peak (90002/90043, 90024?, 90059). This would leave two to four samples in the 7Ma peak (90024?, 90069, 90072, 90082?). 36Cl data and terrestrial ages derived from these values are presented in Table 2. The ages are calculated with a saturation activity of 36Cl in the metal phase of chondrites of (22.1 +/- 2.7) dpm/kg (2 sigma, unpublished). This current best estimate is based on depth profiles in Knyahinya and St. Severin [11] as well as 24 analyses of meteorites with known terrestrial age. The terrestrial age histogram of the Frontier Mountain samples is shown in Fig. 1. The two specimens recovered on Meteorite Moraine show the highest values of roughly 200,000 and 100,000 years, respectively, whereas all other samples have been on Earth for probably less than 100,000 years. It can reasonably be expected from the ice flow pattern near Frontier Mountain that meteorites in the Moraine have larger terrestrial ages than those on the blue ice. Additional samples from the Moraine will be analyzed to verify this difference. All 13 Frontier Mountain meteorites analyzed so far for 36Cl have terrestrial ages below 260,000 years. The age distribution in Fig. 1 is thus similar to those usually found in Antarctica [1-3]. An exception is the Allan Hills Main Icefield, for which the terrestrial ages are shown in Fig. 1 for comparison [3]. We conclude that Frontier Mountain meteorites on average have lower terrestrial ages than those from the Allan Hills Main Icefield. Acknowledgements: We thank EUROMET for the samples used in this study as well as NASA, NSF, and the Swiss National Science Foundation for financial support. References: 1: Nishiizumi K. et al. (1989) Earth Planet. Sci. Lett. 93, 299-313. 2: Schultz L. (1986) Proc. Tenth Symp. Antarctic Meteorites, NIPR Tokyo, 319-327. 3: Nishiizumi K. et al. (1994) Workshop on Meteorites from Hot and Cold Deserts, this volume. 4: Delisle G. et al. (1989) Geol. Jb. E38, 483-513. 5: Delisle G. et al. (1993) Meteoritics 28, 126-129. 6: Graf Th. et al. (1990) Geochim. Cosmochim. Acta 54, 2521-2534. 7: Eugster O. (1988) Geochim. Cosmochim. Acta 52, 1649-1662. 8: Nishiizumi K. et al. (1980) Earth Planet. Sci. Lett. 50, 156-170. 9: Anders E. (1964) Space Sci. Rev. 3, 583-714. 10: Marti K. and Graf Th. (1992) Annu. Rev. Earth Planet. Sci. 20, 221-243. 11: Nishiizumi K. et al. (1989) Proc. Lunar Planet. Sci. Conf. 19th, 305-312 Table 2: 36Cl in metal and terrestrial ages. FRO... mass 36Cl Terrestrial Age [mg] [dpm/kg]^1 age^2 [ka] 90001 85 21.42 +/- 0.24 15 + 55/-15 90002 82 19.93 +/- 0.23 45 + 55/-45 90024 83 14.15 +/- 0.26 195 +/- 65 90043 108 18.90 +/- 0.21 70 +/- 0 90048 109 20.76 +/- 0.72 30 + 65/-30 90050 49 20.30 +/- 0.22 35 + 55/-35 90059 100 19.59 +/- 0.24 55 +/- 55 90069 86 23.68 +/- 0.34 <65 90072 113 22.89 +/- 0.32 <40 90073 116 20.63 +/- 0.29 30 + 55/-30 90082 52 17.33 +/- 0.28 10 +/- 60 90152 89 20.69 +/- 0.37 30 + 60/-30 84033 97 17.19 +/- 1.50 110 +/- 90 1).1-sigma error. 2).Terrestrial age calculated with saturation activity of 36Cl (metal) = (22.1 +/- 2.7) dpm/kg (2 sigma). 3).Data from [1]. Fig. 1: Histogram of terrestrial ages of H-chondrites from Frontier Mountain (this work, Table 2). The age distribution for meteorites from the Allan Hills Main and Near Western Icefields [3] is also shown for comparison. Wlotzka F.* Jull A. J. T. Donahue D. J. 14C Terrestrial Ages of Meteorites from Acfer, Algeria F. Wlotzka1, A. J. T. Jull2, and D. J. Donahue2 1 Max-Planck-Institut für Chemie, Abteilung Kosmochemie, 55020 Mainz, Germany. 2 NSF Arizona AMS Facility, University of Arizona, Tucson, AZ 85721, USA.. The Reg el Acfer in Algeria is a uniform stony desert area of about 30 x 100 km. More than 300 meteorites were collected here in the last years [1]. Because of the comparatively small collection area with uniform climatic and soil conditions, storage conditions for meteorites found here are very similar. The meteorite population thus offers a unique possibility to study relations between terrestrial age and weathering, with the ultimate goal to estimate the total meteorite influx. The terrestrial age was determined by the 14C-method. Small samples of 0.1 to 0.5g can be 14C-dated using accelerator mass spectrometry [2-4]. Jull et al. [2] and Reedy et al. [5] used Knyahinya, Bruderheim and some other chondrites to establish a saturated activity reference level. The production rate of 14C varies with depth and the preatmospheric size of the meteoroid [5]. The precision of terrestrial age determinations is limited by these effects. 14C is of particular interest in warmer climatic regions, where the storage time before a meteorite weathers away is less than at many locations in Antarctica [3,6,7]. 14C-ages have been obtained for meteorites from the semiarid high plains of the southwestern USA [3,8] and arid regions such as the western Libyan desert [6] and Western Australia [9]. The mean residence time of meteorites at such locations is well over 10,000 yr. We have studied the 14C-age distribution of 53 meteorites from Algeria, 51 from the Acfer area (Table 1). Figure 1 presents the 14C-age distribution of these samples, it shows a simple exponential dependence of terrestrial age versus time. The oldest meteorites are at the limit of 14C-dating. A similar age relationship is observed for Western Australian meteorites (Ref. [9], see their Fig. 1 for a comparison with the Acfer data), but with fewer very old samples in the Australian case. This is different from results seen for the southwestern USA, especially Roosevelt County, New Mecico [3,8], where a higher proportion of old meteorites is found (Fig. 1). The less arid and colder high plains of Texas and New Mexico may be more conducive to storage of meteorites for longer periods of time than these areas (which is also shown by a comparison of terrestrial ages and weathering grades, as discussed below), but also we believe that older meteorites may be stored in Pleistocene sediments at these sites, giving a deficit of "young" meteorites, as discussed by Jull et al. [9]. For the Acfer finds the weathering grade (WG) was determined in thin sections using the scheme of Wlotzka [10]. These grades register the degree of alteration of the meteorite minerals as a measure of the weathering intensity, see footnote to Table 1. For Acfer, we observe a dependence of WG on the 14C-age of the meteorite (Fig. 2), although the scatter of the data points is rather large. This shows that besides the residence time on the surface, individual properties of the meteorites, such as cracks and texture must play a role. The highest ages in a given WG are found for meteorites of type L6 or L5, which suggests that a low metal content and recrystallized structure can delay the weathering effects. The opposite effect is seen in the E4 chondrite Acfer 287: It has already reached W4 in only 2.9 Kyr. A better correlation is obtained for the median age versus WG (Fig. 2). A similar correlation of WG with terrestrial age was first observed by us in Roosevelt County meteorites [8]. However, the median terrestrial age for a given Weathering Grade is longer in Roosevelt County than in Acfer, at least for higher WGs (Table 2): Table 2: Median terrestrial age (Kyr) for different WG of Acfer and Roosevelt Co. meteorites Weathering grade: W1 W2 W3 W4 W5 W6 Acfer, Algeria 3 8 13 17 20 n.p. Roosevelt Co., New Mexico n.p. 8 22 27 35 40 n.p. = not present We have also studied the carbonates released by acid etching of the weathering products. We hope to relate trends in the amount of carbonate, delta 13C and 14C-age of the weathering products with the time of weathering and climatic effects. References: [1] Bischoff A. and Geiger T. (1994), Meteoritics, submitted. [2] Jull A. J. T. et al. (1989), GCA, 53, 2095. [3] Jull A. J. T. et al. (1993), Meteoritics, 28, 189. [4] Beukens R.P. et al. (1988), Proc. NIPR Symp. Ant. Met., 1, 224. [5] Reedy R.C. et al. (1993), LPSC XXIV, 1195. [6] Jull A. J. T. et al. (1990), GCA 54, 2895. [7] Boeckl, R. P. (1972) Nature, 236, 25 [8] Jull A.J. T. et al. (1991) LPSC XXII, 665. [9] Jull A. J. T., Bevan A.W.R., Cielaszyk E. and Donahue D.J., this conference. [10] Wlotzka F. (1993), Meteoritics 28, 460. Table 1: 14C terrestrial ages of Algerian meteorites Sample Type WG+ 14C Terrestrial age (dpm/kg) (Kyr) +/- 1.3 El Djouf 003 L6 W41 11.8 +/- 0.15 12.1 Ilafegh 011 L5 W31 6.3 +/- 0.1 17.3 Acfer 019 L6 W31 4.7 +/- 0.1 19.8 Acfer 047 L4 W2 18.5 +/- 0.2 8.4 Acfer 074 L6 W5 3.9 +/- 0.1 21.2 Acfer 111 H3 W1 35.2 +/- 0.3 2.3 Acfer 194 H5 W2 30.4 +/- 0.2 3.5 Acfer 195 H6 W1 48.8 +/- 0.8 recent fall Acfer 197 L6 W2 17.3 +/- 0.2 9.0 Acfer 198 H6 W1 37.5 +/- 0.2 1.8 Acfer 199 L4 W3 24.6 +/- 0.2 6.0 Acfer 200 H3-6 W1 27.5 +/- 0.3 4.3 Acfer 201 L5 W2 10.4 +/- 0.1 13.2 Acfer 203 H5 W3 32.8 +/- 0.2 2.9 Acfer 204 H3 W2 12.3 +/- 0.3 10.9 Acfer 205 H5 W2 11.2 +/- 0.2 11.8 Acfer 206 H5 W1 46.1 +/- 0.6 recent fall Acfer 208 LL5/6 W31 42.2 +/- 0.2 2.2 42.1 +/- 0.3 2.2 Acfer 210 H3 W31 2.67 +/- 0.08 23.6 Acfer 212 H5 W5 8.6 +/- 0.1 13.9 9.7 +/- 0.7 13.0 Acfer 215 L5 W1 4.8 +/- 0.1 19.5 Acfer 216 L6 W2 18.2 +/- 1.3 8.5 * Acfer 218 L6 W5 5.45 +/- 0.53 18.5 Acfer 219 H6 W1 13.3 +/- 0.2 10.4 Acfer 221 H5-6 W2 17.2 +/- 0.4 8.2 Acfer 223 LL6 W3 8.3 +/- 0.1 15.6 Acfer 225 H3 W2 20.2 +/- 0.5 6.9 Acfer 226 H5 W2 16.2 +/- 0.1 8.7 residue 20.0 +/- 0.2 8.4 Acfer 227 H6 W1 26.0 +/- 0.6 4.8 residue 22.3 +/- 0.3 6.0 Acfer 229 H5 W4 4.8 +/- 0.4 18.8 Acfer 236 LL6 W21 41.7 +/- 0.2 2.3 Acfer 244 H5-6 W1 34.5 +/- 0.2 2.5 33.9 +/- 0.3 2.6 Acfer 249 H5-6 W3 23.5 +/- 0.5 5.6 Acfer 263 L6 W1 31.3 +/- 0.4 4.0 33.7 +/- 0.3 3.4 Acfer 274 H6 W3 2.8 +/- 0.1 23.2 Acfer 276 H5 W3 7.7 +/- 0.1 14.8 Acfer 280 L6 W4 10.4 +/- 0.1 13.1 Acfer 281 L5 W4 10.2 +/- 0.1 13.3 * Acfer 287 E4 W41 29.0 +/- 0.3 2.9 Acfer 288 L5 W3 7.1 +/- 0.1 16.3 Acfer 290 L6 W3 1.5 +/- 0.1 29. Acfer 291 H6 W2 16.6 +/- 0.1 8.5 Acfer 292 H5 W2 18.5 +/- 1.3 7.6 Acfer 296 LL5-6 W21 15.4 +/- 0.1 10.6 Acfer 298 L6 W5 <0.30 >42.4 Acfer 302 H5 W3 16.1 +/- 0.2 8.8 Acfer 303 H5-6 W3 13.3 +/- 0.2 10.3 Acfer 305 L6 W5 7.6 +/- 0.6 15.1 +/- 1.5 Acfer 306 L6 W3 18.5 +/- 10.4 8.4 Acfer 307 L5 W2-5 1.0 +/- 0.3 31.5 +/- 3 Acfer 309 L6 W3 3.6 +/- 0.5 22.0 +/- 1.8 Acfer 310 H6 W3 11.9 +/- 0.4 11.3 Acfer 315 L/LL6 W4 0.7 +/- 0.5 35.5 +/- 5.5 * Age may be too low due to 14C contamination by weathering products. + WG: weathering grades; W1 minor, W2 moderate, W3 heavy, W4 total oxidation of metal and troilite, W5 beginning, W6 massive alteration of silicates. 1 Weathering grade from [1] Benoit P. H.* Sears D. W. G. Terrestrial Age Clustering of Meteorite Finds from Sites in Antarctica and Hot Deserts Many concentrations of meteorite finds have been found in hot and cold desert regions of the world. The sites include the deserts of northern Africa, the western United States, and Australia, in addition to the ice fields of Antarctica [1,2,3]. These large groups of meteorite finds allow detailed study of the nature of the meteorite flux in the pre- historical past [e.g., 4,5]. In this paper we discuss terrestrial age data, derived from cosmogenic radionuclide abundances (such as 14C and 36Cl) and natural thermoluminescence data for meteorite finds from sites in both hot and cold desert sites. We find that at all the sites we have examined to date there is distinct clustering of either terrestrial ages or surface exposure ages of the meteorite finds; we suggest that meteorite concentration in hot and cold deserts is generally episodic, reflecting changes in regional climate or changes in local conditions such as directions of ice or stream flow over time. The natural thermoluminescence (TL) level of a meteorite find is determined by radiation dose, temperature, and time. At a given temperature, the equilibrium natural TL level of a meteorite is directly proportional to the dose rate. Since the radiation dose rate for meteorite finds is inevitably far less than during irradiation in space, the natural TL level of a meteorite on Earth will decrease over time until a new, much lower equilibrium level is reached, this level being determined by terrestrial temperature and radiation dose. At equilibrium, higher temperatures result in lower natural TL levels. Under non-equilibrium conditions, the rate of TL decay is determined by temperature. The systematics of natural TL decay in ordinary chondrites have been detailed by theoretical calculations and laboratory heating experiments [6] and comparisons have been made with terrestrial age estimates obtained using cosmogenic radionuclide abundances [7,8]. However, the latter comparison makes the assumption that meteorite finds have one-stage terrestrial thermal histories which is not necessarily the case for Antarctic meteorites. In fact, TL results can be particularly interesting in cases where the meteorites have experienced discrete multi-stage terrestrial thermal histories. The TL methodology and data reduction techniques have been discussed elsewhere [e.g., 8,9]. Depth effects, which are common problems in cosmogenic nuclide studies, are not a significant factor in governing natural TL levels except for the largest meteorites [10]. In the present study we limit our database to ordinary chondrites of type 3.7 and higher which are thought to have the same major TL phosphor. It is possible to compared TL data for achondrites with those of ordinary chondrites but corrections must be made for "anomalous fading" [9]. Roosevelt County/western United States. In Fig. 1 we compare our TL data with the 14C-derived terrestrial ages for meteorite finds from the western United States. We show in Fig. 1a the 1980 version of this plot, where 14C terrestrial ages were obtained by Boeckl by beta counting [11] and in Fig. 1b the current version where 14C terrestrial ages are from Jull's AMS data [12]. Apparently between 1980 and 1990 TL data was providing more reliable information on meteorite terrestrial ages than 14C data and there was a misplaced confidence in 14C-derived terrestrial ages. Natural TL decay curves for various "storage temperatures" are also shown on Fig. 1b. In general, the meteorites plot close to the 20 degrees C TL decay line. From data for finds from the western United States (Fig. 1b) it is apparent that most of these meteorites have fairly short terrestrial ages, with most having terrestrial ages <5 ka. There are few meteorites with terrestrial ages between about 5 and 30 ka, and even fewer with ages >30 ka. This distribution is very different from that of the Roosevelt County meteorite finds (Fig. 2, 14C data from [13]) in which only a very few meteorites have terrestrial ages <5 ka and all other meteorites have terrestrial ages >20 ka. In both databases there are a few meteorites (Ladder Creek and RC-064) with TL levels far lower than that expected in consideration of their 14C-derived terrestrial ages; these meteorites were probably reheated in space prior to Earth impact [14]. The difference in age distributions between these two collections almost certainly reflects differences in collection and recovery history. The Roosevelt County samples are from highly localized blow-outs which were systematically searched, while the western United States finds are chance discoveries over an ~800 km region. However, it is interesting that in both databases there is a fairly distinct gap between about 10 - 20 ka in which few meteorites are found. Jull et al. [12] suggested that the Roosevelt County distribution was produced by the loss of meteorites in the intermediate age range during the blow-out events. This interpretation seems reasonable because of the small size of these meteorites (generally <100 grams). However, the western U.S. finds are generally >2 kg and occasionally >100 kg. The only meteorite which has an age in the 10-20 ka range is the L6 chondrite Bluff which is especially large (142 kg). One possible conclusion from this is that between about 10 to 20 ka ago conditions were unsuitable for meteorite accumulation/preservation over most of the western United States and that a regional variable, such as an unfavorable climate, was responsible for loss of most meteorites in the gap. North Africa. A plot of 14C- derived terrestrial ages [2] vs. natural TL data for a fairly small group of finds from Daraj, Libya looks similar to the equivalent plot for finds from the western United States (Fig. 1b), although the meteorites tend to plot along a TL decay curve for a storage temperature of 30 degrees C instead of 20 degrees C (see [8]). If we plot our TL data for meteorite finds from the Acfer, Hammadah al Hamra, Ilafegh, Reggane, and Tanezrouft sites along this curve (14C data are not available for these meteorites) we find that there are four distinct groups (Fig. 3). Three of the groups appear to reflect the terrestrial age distribution, whereas the fourth probably consists of reheated meteorites, which most likely have had their TL drained in small perihelia orbits. North Africa has had a complex climate history over the last 15 ka, but it appears that the area was much wetter in the past and fluvial processes were especially active at about 10,000 and 5100 BP, at least in some portions of the current Sahara desert [15]. It thus appears that, as suggested by Jull et al. [2], meteorite accumulation/preservation in this region is episodic and controlled by climate variations. Antarctica. Meteorites in Antarctica can be in one of two thermal states, (1) buried in the ice or (2) exposed on the ice surface. When on the surface of the ice meteorites can have temperatures higher than the air as a result of solar heating [16]. Calculations indicate that the rates of TL decay at temperatures experienced while encased in the ice are insignificant compared to those at surface temperatures. Thus, the natural TL level of an Antarctic meteorite find is largely determined by the "surface exposure age" rather than the total terrestrial age. We have calculated surface exposure ages for a group of meteorites for which terrestrial ages, largely determined from 36Cl abundances [17] are available (Fig. 4). To make this calculation we assume an average surface exposure temperature of -15 degrees C [16] and an initial TL level of 100 krad [18]. Ignoring the group of meteorites with apparently small terrestrial ages but very large surface exposure ages because these are "reheated meteorites" which are also observed in every other database (i.e., Figs 1b and 2), there is a large range of exposure ages. Most meteorites spent <=50% of their terrestrial histories exposed on the ice surface. There appears to be a hiatus in exposure ages between about 0.18 and 0.2 Ma, and a "ceiling" at about 0.25 Ma with even meteorites with terrestrial ages ~1 Ma having surface exposure ages of ~0.2 Ma. Either 0.2 Ma is the length of time required to move a meteorite across a blue ice field and back onto active ice or an event cleared the icefield of meteorites with surface exposure ages >0.25 Ma. One could also interpret the meteorites with small surface exposure ages and small terrestrial ages as locally derived, while those with greater surface exposure ages and terrestrial ages could represent meteorites derived from some distance away and transported to the field by the ice. In this case, our data agree with the suggestions of Huss [19] in that it appears that locally derived meteorites dominate the dataset. The data shown in Fig. 4 are dominated by meteorites from the Allan Hills Main and Near Western fields. We have previously suggested on the basis of natural TL data that there are differences in surface exposure ages between fields and, in a few cases, in regions within fields. Among our observations, we have noted that there are apparent differences in surface exposure ages between the Elephant Moraine icefields and the Allan Hills Main icefield, with the Elephant Moraine meteorites generally have fairly small surface exposure ages [18]. We have also noted that there are differences between the Lower and Upper Ice Tongues at the Lewis Cliff site, with meteorites from the Lower Ice Tongue having apparently lower surface exposure ages than those with the Upper Ice Tongue [20]. The cosmogenic radionuclide database for Lewis Cliff meteorites is still very small, but two meteorites from the Lower Ice Tongue plot with the group with small terrestrial ages and small surface exposure ages whereas a single Upper Ice Tongue meteorite plots with the meteorites with high surface exposure age. Conclusions. We have used natural TL data to determine terrestrial ages and, in the case of Antarctic meteorites, surface exposure ages. We find that there is evidence that meteorite accumulation/preservation has been episodic in the western United States and the Sahara desert. This presents difficulties in estimating meteorite flux from numbers of meteorites on accumulation surfaces (e.g., [12]). We also find that Antarctic meteorite accumulation surfaces may also be episodic in activity and that some fields are more "stable" than others. Acknowledgements. We wish to thank Tim Jull for sharing data and samples for meteorite finds from the western U.S. and the Sahara. We also thank Steve McKeever, Kuni Nishiizumi and Bill Cassidy for discussions and the Meteorite Working Group of NASA, Marilyn Lindstrom and Robbie Score for Antarctic meteorite samples. This study supported by NASA grant NAG 9-81 and NSF grant 9115521. References. [1] Huss G.I. and Wilson I.E. (1973) Meteoritics 8, 287. [2] Jull A.J.T., Wlotzka F., Palme H., and Donahue D.J. (1990) Geochim. Cosmochim.Acta 54, 2895 [3] Cassidy W.A., Harvey R., Schutt J., Delisle G., and Yanai K. (1992) Meteoritics 27, 490. [4] Benoit P.H. and Sears D.W.G. (1992) Science 255, 1685. [5] Lipschutz M.E. and Samuels S.M. (1989) Geochim. Cosmochim. Acta 55, 19. [6] McKeever S.W.S. (1982) Earth Planet. Sci. Lett. 58, 419. [7] Sears D.W.G., Hasan F.A., Myers B.M., and Sears H. (1989) Earth Planet. Sci. Lett. 99, 380. [8] Benoit P.H., Jull A.J.T., McKeever S.W.S., and Sears D.W.G. (1993) Meteoritics 28, 196. [9] Sears D.W.G., Benoit P.H., Sears H., Batchelor J.D., and Symes S. (1991) Geochim. Cosmochim. Acta 55, 3167. [10] Benoit P.H., Chen Y., and Sears D.W.G. (1994) Lunar Planet. Sci. 25, 99. [11] Boeckl R.S. (1972) Nature 236, 25. [12] Jull A.J.T., Donahue D.J. and Wlotzka F. (1993) Meteoritics 28, 188. [13] Jull A.J.T., Wlotzka F., and Donahue D.J. (1991) Lunar Planet. Sci. 22, 667. [14] Benoit P.H., Sears D.W.G., and McKeever S.W.S. (1991) Icarus 94, 311. [15] Pachur H.J. (1980) In The Geology of Libya, p. 781-788. [16] Schultz L. (1990) In LPI Tech. Rpt. 90-03, Workshop on Antarctic Meteorite Stranding Surfaces, pp. 56- 59. [17] Nishiizumi K., Elmore D., and Kubik P.W. (1989) Earth Planet. Sci. Lett. 93, 299. [18] Benoit P.H., Roth J., Sears H., and Sears D.W.G. (1994) J. Geophys. Res. 99, 2073. [19] Huss G.R. (1990) Meteoritics 25, 41. [20] Benoit P.H., Sears H., and Sears D.W.G. (1992) J. Geophys. Res. 97, 4629. Figure 1, showing Western United States' finds, appears here in the hard copy. Figure 2, showing Roosevelt County meteorites, appears here in the hard copy. Figure 3, showing Saharan Meteorites, appears here in the hard copy. Figure 4, showing Antarctic ordinary chondrites, appears here in the hard copy. Thursday, July 21, 1994 DIFFERENT POPULATIONS? 3:30 - 5:00 p.m. Chair(s): G. Crozaz Wolf S. F.* Lipschutz M. E. Yes, Meteorite Populations Reaching the Earth Change with Time! The answer provided by the title of this Workshop contribution is based upon data obtained by radiochemical neutron activation analysis (RNAA) and accelerator mass spectrometry (AMS). We have chosen H4-6 chondrites as a paradigm. The RNAA measurements involve volatile trace and ultratrace elements (whose concentrations reflect the meteorites' preterrestrial thermal histories): the AMS results for cosmogenic 300 ky 36Cl (and other longer-lived cosmogenic radionuclides) provide nominal terrestrial ages for Antarctic meteorites. Time-dependent source changes of the meteorite flux in Earth -- which are not predicted by existing statistical, Monte Carlo models for meteoroid dynamics -- can be short-term and/or long-term. To assess these, we compare compositions of H chondrite sample suites chosen by physical criteria that are independent of the meteorites' compositions. In each H chondrite, we measure 10 volatile (Rb, Ag, Se, Cs, Te, Zn, Cd, Bi, Tl, In) and a few other trace elements by RNAA. Since suite- differences are subtle, we use well-established multivariate statistical techniques of linear discriminant analysis and logistic regression to classify the two sample suites compositionally and test the model- dependent level at which we can disprove the Null hypothesis that the two sample suites derive from the same parent population. We also test this Null hypothesis using a novel, model-independent approach -- randomization-simulation [1,2] -- which provides confidence levels at which we can assert that the suites being compared are compositionally distinguishable. Short-Term Variations From their circumstances of fall, Dodd et al. [3] determined that a significant elongate cluster of 17 co-orbital H chondrite falls (H Cluster 1) in May, 1855-1895, records encounters with two or three closely-spaced and probably related meteoroid stream components, each of which was met near perihelion. Meteorites of H Cluster 1 vary widely in petrographic type (3-6), shock facies (a-d) and 21Ne exposure age (<5-50 My). However, RNAA studies reveal that the volatile trace element signature of 13 of them is distinct from that of 45 other H4-6 chondrite falls not of H Cluster 1 (cf. Fig. 1). Hence, members of H Cluster 1 have a common thermal history and, thus, derive from a common source region in an H chondrite parent body. Similar studies are about to commence of additional H4-6 chondrite falls suspected to be members of other co-orbital streams, to test how general it is that H chondrites -- like H Cluster 1 -- selected by one criterion (fall parameters) prove by another criterion (contents of volatile trace elements) to be significantly distinguishable from other H chondrite falls. In this connection, it is important to note that the fall period for H Cluster 1 chondrites (days 133-147) coincide with days on which the two known asteroids that approach the Earth most closely, do so: 1993 KA2 - 0.0010 AU on day 140 (20 May); and 1993 KA - 0.0071 AU on day 137 (17 May) [3]. The latter is of particular interest since it is one of the surprisingly numerous asteroids that are co-orbital in an Earth-like orbit [4]. Long-Term Variations To examine whether variations in the meteoroid flux to Earth occur on the ky time scale it is necessary to make measurements on meteorites found in Antarctica which have terrestrial ages ranging up to 1 My. Meteorites recovered from Victoria Land and from Queen Maud Land, Antarctica, have different terrestrial age distributions and mean terrestrial ages, 300 and 100 ky, respectively [5]. There is no doubt that meteorites of rare types discovered in Antarctica differ markedly from those falling more recently elsewhere on Earth. Whether such Antarctic/non-Antarctic differences of a pre-terrestrial origin exist is more controversial. It has been demonstrated that Antarctic H4-6 (and L4-6) chondrites mainly from Victoria Land are, on average, compositionally distinguishable from falls [1, 6, 7]. A significant difference between these suites is also detected by thermoluminescence [8] but not by noble gas measurements [9]. Some have argued that differences in contents of volatile trace elements in Antarctic and non-Antarctic H4-6 chondrites, for example, reflect weathering in Antarctica and/or are artifacts of the data treatment. These possibilities are discussed in detail elsewhere [10] and found not to be viable. Significant compositional differences exist between H4-6 chondrite suites from Victoria Land (34 samples) and Queen Maud Land (25 samples), Antarctica, presumably reflecting mean terrestrial age differences [10]. This time-dependent compositional difference has been established directly by comparing the sample suite of 58 H4-6 chondrite falls with Antarctic suites having specific nominal terrestrial age ranges [11]. Compositionally, 13 Antarctic H4-6 chondrites with nominal terrestrial ages <=50 ky are not significantly distinguishable from 58 falls, i.e., are consistent with the Null hypothesis that these sample suites are drawn from the same parent population. However, very different results are obtained when 12 Antarctic chondrites with nominal terrestrial ages of 50-70 ky or when 13 Antarctic chondrites having nominal terrestrial ages >=70 ky are studied: each differs compositionally from 58 falls. When these two Antarctic suites are combined, the compositional distinction between 58 falls and the 25 samples with nominal terrestrial ages >50 ky is -- like H Cluster 1 -- significant [11] beyond any reasonable doubt (Fig. 2). As a consequence of this time-dependent change, 34 Victoria Land H4-6 chondrites differ compositionally, on average, from 58 falls while 25 from Queen Maud Land do not, i.e., are consistent with the Null hypothesis that the Queen Maud Land suite and the suite of falls are drawn from the same parent population [12]. Hence, both on the short-term and the long-term, there is ample evidence that the Earth's sampling of meteoritic matter has varied with time. How this occurred dynamically is now being considered [4,13]. (This research was sponsored by: NASA grants NAG 9-48, NAG 9-580, NAGW- 3396 and NAGW-3460; NSF grants EAR-8916667, EAR-9214636 and EAR-9305859; DOE grant DE-FG-07-80ER1 0725 J; and NATO grant 0252/89.) References: [1] Lipschutz M. E. and Samuels S. M. (1991) Geochim. Cosmochim. Acta 55, 19-47. [2] Wolf S. F. and Lipschutz, M. E. (1994) Advances in Analytical Geochemistry (M. Hyman and M. Rowe, eds.) in press. [3] Dodd R. T., Wolf S. F. and Lipschutz M. E. (1993) J. Geophys. Res.-Planets 98, 15105-15118. [4] Rabinowitz D. L., Gehrels T., Scotti J. V., McMillan R. S., Perry M. L., Wisniewski W., Larson S. M., Howell E. S., and Mueller B. E. A. (1993) Nature 363, 704-706. [5] Nishiizumi K. (1987) Nucl. Tracks Radiat. Meas. 13, 209-273. [6] Dennison J. E. and Lipschutz M. E. (1987) Geochim. Cosmochim. Acta 51, 741-754. [7] Kaczaral P. J., Dodd R. T. and Lipschutz M. E. (1989) Geochim. Cosmochim. Acta 53, 491-501. [8] Benoit P. H. and Sears D. W. G. (1993) Icarus 101, 188-200. [9] Schultz L., Weber H. and Begemann F. (1991) Geochim. Cosmochim. Acta 55, 101-125. [10] Wolf S. F. and Lipschutz M. E. (1993) J. Geophys. Res.-Planets, submitted. [11] Michlovich E. S., Wolf S. F., Wang M.-S., Vogt S., Elmore D. and Lipschutz M. E. (1993) J. Geophys. Res.-Planets, submitted. [12] Wolf S. F. and Lipschutz M. E. (1993) J. Geophys. Res.-Planets, submitted. [13] Hughes D. (1993) WGN, J. Intern. Meteor Org. 21, 254-258. [14] Lingner D. W., Huston T. J., Hutson M. and Lipschutz M. E. (1987) Geochim. Cosmochim. Acta 51, 727- 739. Fig. 1. Logistic regression based upon data for 10 volatile trace elements reveals perfect separation of 13 H Cluster 1 falls from 45 other H4-6 chondrite falls, hence compelling evidence of a compositional difference. Sources of data: filled symbols [3]; open symbols [14]. The compositional uniqueness of H Cluster 1 chondrites from all other falls signals that the Cluster derives from a unique parent region. Fig. 2. Logistic regression based upon data for 10 volatile trace elements reveals very good separation of 25 Antarctic H4-6 chondrites (nominal terrestrial ages >50 ky) from 58 falls [11]. By model-dependent or -independent methods, p values <0.001 signal very strong evidence that these Antarctic H4-6 chondrites derive from source regions different than the one(s) that yielded contemporary H4-6 chondrite falls. Wolf S. F.* Lipschutz M. E. Applying the Bootstrap to Antarctic and Non-Antarctic H Chondrite Volatile Trace Element Data Recently developed bootstrap statistical methods are making a significant impact in the field of chemometrics. Bootstrap methods substitute iterative calculations for theoretical approximations used in many classical statistical techniques (1). Randomization-simulation (2) illustrates the application of the bootstrap method to the classical statistical techniques linear discriminate analysis and logistic regression as they are used to compare populations of Antarctic H and L chondrites to their non-Antarctic counterparts on the basis of volatile trace element composition. These techniques test the Null hypothesis of no difference in volatile trace element composition between the two populations and calculates model-independent significance levels. Volatile trace element composition was selected for these tests because of their sensitivity to subtle thermal processes. By utilizing both linear discriminate analysis and randomization-simulation, two different testing approaches, we can assess the accuracy of model-dependent significance levels and determine whether discrimination is the result of invalid distributional assumptions, too many independent variables or too few samples. In this study we examine the suggestion that the flux of H chondrite material has changed with respect to time (3). A variety of differences have been observed between Antarctic and non-Antarctic meteorite populations (4 and references therein). Comparisons of samples from Victoria Land and Queen Maud Land show differences in labile trace element contents (5) and induced thermoluminescence parameters (6). H chondrite meteorite populations from Victoria Land, Antarctica; Queen Maud Land, Antarctica; and modern falls have average terrestrial ages of 300 ky, 100 ky, and < 200 y, respectively (7). Here we compare each of these three populations to each other on the basis of the volatile trace elemental composition using multivariate linear discriminant analysis. Volatile trace element data exists for 117 H4-6 chondrites. The data set includes: Rb, Ag, Se, Cs, Te, Zn, Cd, Bi, Tl, and In (listed in order of increasing volatility). This data set consists of 34 samples from Victoria Land, Antarctica; 25 samples from Queen Maud Land, Antarctica; and 58 modern falls. Multivariate linear discriminant analysis based on the concentrations of ten volatile trace elements, demonstrates that the terrestrial collection of H4-6 chondrites contains compositionally distinct subpopulations. Comparison between Antarctic and non-Antarctic, Victoria Land and non-Antarctic, Victoria Land and Queen Maud Land, Queen Maud Land and non-Antarctic H4-6 chondrites by linear discriminant analysis all reveal significant differences in volatile trace element composition (Table 1). Model-independent p-values, calculated by linear discriminant analysis based randomization-simulation, differ to varying degrees the with model-dependent p-values. However, significant difference in volatile trace element composition is indicated. Both multivariate linear discriminant analysis and randomization- simulation provide extremely strong evidence of a difference in volatile trace element composition between these three populations. These techniques do not give the cause or causes of the observed differences. Insights to the causes of difference might be gained by determining the contribution that each element has to discrimination. A method that allows the visualization of each element's contribution to discrimination is a canonical details plot (Figure 1). This figure graphically illustrates the centroid of each population in two- dimensional canonical space. The centroids of each population appear with a circle corresponding to the 95% confidence region (8). The elemental rays indicate the loading that these variables have on each dimension in this test space. Figure 1 suggests that all 10 volatile trace elements affect discrimination between these populations to varying degrees. However, the loading that each element has depends on which populations are being compared. Elements that load the highest when Antarctic (Victoria Land and Queen Maud Land) and non-Antarctic H4- 6 chondrites are compared are Ag, Cd, and Tl. Rb, Cs, and Zn also load to a lesser degree. Discrimination between Victoria Land and Queen Maud Land H4-6 chondrite populations is dominated by Se, Cs, Te, and Bi. Discrimination between Queen Maud Land and non-Antarctic H4-6 chondrite populations is dominated by Ag, Cd, Tl, and In. Table 1. Results of linear discriminant analysis and randomization- simulation based on the concentrations of 10 volatile trace elements. population number of samples model-dependent model-independent p-value p-value Antarctic 59 <0.0001 <0.001 non-Antarctic 58 Victoria Land 34 0.0178 0.065 Queen Maud Land 25 Victoria Land 34 <0.0001 <0.001 non-Antarctic 58 Queen Maud Land 25 0.0002 0.002 non-Antarctic 58 Figure 1. Canonical details plot of volatile trace element composition of 34 Victoria Land, 25 Queen Maud Land, and 58 modern falls. Possible causes of difference in volatile trace element composition include trivial causes such as sampling bias, analytical bias, pairing, and weathering. Similar differences in volatile trace element composition are revealed when canonical details of samples from a single analyst are plotted. Differences in volatile trace element composition are, therefore, inconsistent with sampling or analytical bias. Noble gas and cosmogenic nuclide data indicate that pairing of Antarctic samples is essentially nonexistent in out database (9&10). All Victoria Land samples are of weathering type A, A/B, or B. Samples from Queen Maud Land were selected on the basis of minimal visible oxidation in the hand specimens. Differences in volatile trace element composition shown in Figure 1 are inconsistent with leaching resulting from weathering. Differences observed between Victoria Land, Queen Maud Land, and modern falls are consistent with a temporally varying sampling of H4-6 chondrite material that is heterogeneous with respect to volatile trace elements. Different thermal histories for these three populations are indicated. References: (1) Efron, B. (1982) The Jackknife, the Bootstrap and Other Resampling Plans. CBMS-NSF 38. (2) Lipschutz, M. E. and Samuels, S. M. (1991) Ordinary chondrites: Multivariate statistical analysis of trace element contents. Geochim. Cosmochim. Acta 55, 19-34. (3) Dennison, J. E., Lingner, D. W. and Lipschutz, M. E. (1986) Antarctic and non- Antarctic meteorites form different populations. Nature 319, 390-393. (4) Koeberl, C. and Cassidy, W. A. (1991) Differences between Antarctic and non-Antarctic meteorites: An assessment. Geochim. Cosmochim. Acta 55, 3-18. (5) Wolf, S. F. and Lipschutz, M. E. (1993) Chemical Studies of H Chondrites - IV. New Data and Comparison of Antarctic Populations. J. Geophys. Res. -Planets, submitted. (6) Benoit, P. H. and Sears, D. W. G. (1992) The break-up of the H chondrite parent body and the delivery of fragments to Earth. Lunar Planet. Sci. (abstract) 23, 85-86. (7) Nishiizumi, K. Elmore, D. and Kubik, P. W. (1989) Update on terrestrial ages of Antarctic meteorites. Earth Planet. Sci. Lett. 93 299-313. (8) Mardia, K. V., Kent, J. T. and Bibby, J. M. (1979) Multivariate Analysis, New York: Academic Press. (9) Loeken, T., Scherer, P. and Schultz, L. (1993) Noble gases in twenty Yamato H chondrites: Comparison with Allen Hills chondrites and modern falls. Lunar Planet. Sci., (abstract) 24, 889-890. (10) Michlovich, E. S., Wolf, S. F., Wang, M.- S., Vogt, S., Elmore, D. and Lipschutz, M. E. (1993) Chemical studies of H chondrites- V. Temporal Variation of sources. J. Geophys. Res. - Planets, submitted. Loeken T.* Schultz L. The Noble Gas Record of H Chondrites and Terrestrial Age: No Correlation On the basis of statistically significant concentration differences of some trace elements, it has been suggested that H-chondrites found in Antarctica and Modern Falls represent members of different extraterrestrial populations with different thermal histories (e.g. [1]). It was also concluded that H-chondrites found in Victoria Land (Allan Hills) differ chemically from those found in Queen Maud Land (Yamato Mountains), an effect which could be based on the different terrestrial age distribution of the two groups [2]. This would imply a change of the meteoroid flux hitting the Earth on a timescale that is comparable to typical terrestrial ages of Antarctic chondrites. A comparison of the noble gas record of H-chondrites from the Allan Hills icefields and Modern Falls [3] shows that the distributions of cosmic ray exposure ages and the concentrations of radiogenic 4He and 40Ar are very similar. In an earlier paper [4] we have compared the noble gas measurements of 20 Yamato H-chondrites with meteorites from the Allan Hills region and Modern Falls. Here also very similar distributions were found. A possible variation of the meteoroid flux with time is perhaps not very obvious because the distribution of terrestrial ages of Allan Hills and Yamato Mountain meteorites show a broad overlap. For 37 finds of these groups trace element and noble gas concentrations as well as terrestrial ages are known. In this paper the distribution of cosmic ray exposure ages and radiogenic 4He and 40Ar contents as a function of terrestrial age is investigated. The cosmic ray exposure ages are derived from the concentration of cosmogenic 21Ne using the production rates and shielding corrections given by Eugster [5]. Figure 1 shows in its left part the exposure ages of H-falls calculated from literature values [6]. Only measurements with cosmogenic 22Ne/21Ne between 1.08 and 1.18 were taken into account to reduce uncertainties caused by extreme shielding conditions. The distribution shows the well-known 7-Ma-cluster indicating that about 40 % of the H-chondrites were excavated from their parent body in a single event. The distribution for all Antarctic H-chondrites from [3] and [4] is shown in the middle part. The right part of the figure shows the exposure age of Antarctic H-chondrites plotted as function of their terrestrial age. No correlation between exposure age and terrestrial age is observed. Both populations, Antarctic meteorites and falls, exhibit the same characteristic feature: a major meteoroid producing event about 7 Ma ago. This indicates that one H-group population delivers H-chondrites to Antarctica and the rest of the world. Similar comparisons are made for the radiogenic nuclides 4He and 40Ar. H-falls show a maximum between 1250 X 10^(-8) cm^3 STP/g and 1500 X 10^(-8) cm^3 STP/g of radiogenic 4He. For Antarctic H- chondrites also most samples fall into this region independent of their terrestrial age (Fig. 2). The distribution of the concentration of 40Ar in H-falls has a peak with a maximum between 5500 X 10^(-8) cm^3 STP/g and 6000 X 10^(-8) cm^3 STP/g. This distribution is also found in Antarctic meteorites (Fig. 3). We conclude that cosmic ray exposure ages and thermal history indicators like radiogenic noble gases show no evidence of a change in the H-chondrite meteoroid population during the last 200000 years. References: [1] Dennison J. E. et al. (1986) Nature, 319, 390-393. [2] Wolf S. F. and Lipschutz M. E. (1992) Lunar Planet. Sci, 23, 1545-1546. [3] Schultz L. et al. (1991) Geochim. Cosmochim. Acta, 55, 59-66. [4] Loeken Th., Scherer P. and Schultz L. (1993) Lunar Planet. Sci., 24, 889-890. [5] Eugster O. (1988) Geochim. Cosmochim. Acta, 52, 1649-1662. [6] Schultz L. and Kruse H. (1989) Meteoritics, 24, 155-172; and supplement. Fig. 1: Exposure age distribution of H-chondrites. The histogram of H-falls (left) shows the well-known 7 Ma peak which is also well pronounced for Antarctic H-chondrites (middle). No correlation of exposure age and terrestrial age is observed (right). Fig. 2: Contents of radiogenic 4He in H-chondrites. The distribution for H-chondrites from falls is shown in the left diagram, that for Antarctic meteorites in the middle part. For Antarctic meteorites no dependence from the terrestrial age is observed (right). Fig. 3: Distribution of radiogenic 40Ar in H-chondrites (left: falls; middle: Antarctic meteorites). No correlation between radiogenic 40Ar and terrestrial age is observed (right). Benoit P. H.* Sears D. W. G. The Antarctic Collection and Changes in the Meteorite Flux Over Time: The Lingering Death of a Subgroup of H Chondrites Differences between the Antarctic meteorite find collection and modern falls have been noted since the initial Antarctic expeditions. Among the more obvious differences are the lesser abundance of iron meteorites among Antarctic meteorites compared to modern falls and a tendency for Antarctic meteorite finds to be smaller than their equivalents among non-Antarctic meteorites [1,2]. On a finer scale, differences have been noted in the diversity of types of iron meteorites in the Antarctic collection compared to modern falls [3] and in trace element concentrations between H5 chondrites in the two collections [4]. The significance of some of these differences is uncertain due to the effects of weathering and the largely unknown amount of "pairing" within the Antarctic collection [2,5]. Differences have not been observed in the cosmic ray exposure age distribution for ordinary chondrites [6] and in non-volatile bulk composition [7]. Here we review the induced thermoluminescence (TL) database for H chondrites which, together with metallographic cooling rates, does show a difference between the Antarctic and non-Antarctic meteorite collections and, in fact, indicates that the difference correlates with terrestrial age within the Antarctic collection. We present new data on H6 chondrites, discuss 26Al data for H chondrites, and discuss the implications of these data for changes in the meteorite flux over the last few hundred thousand years. The induced TL data for H5 chondrites from the Allan Hills show two distinct groups, one with TL peak temperatures <190 degrees C and the other with peak temperatures >190 degrees C (Fig. 1) [8,9]. The latter group is significantly different in induced TL properties from H5 chondrites in the modern falls. We have suggested that the difference in TL peak temperatures is the result of different thermal histories; either the meteorites of the >190 degrees C group cooled through the order/disorder transition of feldspar more rapidly than meteorites of the <190 degrees C group, or these meteorites were annealed at high temperatures subsequent to crystallization. Meteorites of the >190 degrees C group have significantly higher metallographic cooling rates, >100 K/Myr compared to about 30 K/Myr for the members of the <190 degrees C group. The >190 degrees C group meteorites share two other properties, namely, they all have cosmic ray exposure ages of about 8 Ma (although not all Antarctic meteorites with 8 Ma exposure ages belong to this group) and they generally have relatively large 3He/21Ne and 22Ne/21Ne ratios, which is indicative of a fairly small size during irradiation in space. Petrographic observations do not indicate that the meteorites of the >190 degrees C group were shocked to a greater degree than meteorites of the <190 degrees C group [9]. Weathering does not explain the >190 degrees C group because (1) hand-specimen descriptions of meteorites of the >190 degrees C group do not indicate that they are more weathered than those of the <190 degrees C group [9], (2) laboratory acid washing experiments indicate that, while weathering does reduced the overall induced TL sensitivity of Antarctic meteorites, it does not change TL peak temperatures or widths [10], (3) the metallographic profiles in these meteorites cannot be modified by weathering, and (4) this behavior is restricted to samples with CRE ages of 8 Ma and there is no reason that weathering would be selective on the basis of cosmic ray exposure age. Furthermore, this group is not caused by "pairing" of a single unusual meteorite. Natural TL and cosmogenic noble gas data [6,9] are not consistent with the distribution of induced TL being due to pairing. Nor are there large numbers of regolith breccias in this group, as was suggested by Kallemyn et al. [7] in their study of bulk chemistry. The >190 degrees C group appears to be correlated with the differences observed in trace element studies of H chondrites [4,11]. We will hereafter refer to the >190 degrees C meteorites as the "rapidly cooled" H chondrites and the <190 degrees C group meteorites as the "normal" H chondrites. We suggested on the basis of data from the TL survey of Antarctic meteorites [9] that the rapidly cooled group included H chondrites of types 4 and 6 in addition to the petrologic type 5 meteorites which dominated the original study. We described H4 chondrites which belonged to this group, using both induced TL data and metallographic cooling rate data [12]. We have recently been examining a collection of Antarctic H6 chondrites, as well as a group of H6 modern falls for comparison. Thus far we have obtained new induced TL data for 12 Antarctic H6 chondrites; metallographic data will be obtained in the near future. Of the twelve Antarctic meteorites, only three, MBRA 76001, META 78019 and ALHA 80126, appear to belong to the rapidly cooled H chondrite group and one of these has a cosmic ray exposure age of about 8 Ma (no CRE age data are available for the other two). As was the case for the H4 and H6 databases, most of the Antarctic H6 chondrites which have normal induced TL peak temperatures have cosmic ray exposure ages greater than 18 Ma. There are two exceptions to this rule, namely ALHA 76008 which has an apparent CRE age of about 2 Ma, but which has been previously documented to have a multi-phase cosmic ray exposure history with a total exposure time of well over 20 Ma [6] and ALHA 81037, which has a CRE age of about 8 Ma but which is apparently a member of the normal H chondrite group. It is thus equivalent to many H chondrites among the modern falls, many of which have cosmic ray exposure ages of about 8 Ma but all of which have normal induced TL and metallographic cooling rates. The discovery of a meteorite like ALHA 81037 in the collection is not unexpected and, in fact, the apparent rarity of such meteorites, as will be discussed below, largely reflects a bias created in most Antarctic - non-Antarctic meteorite comparison studies which tend to use the Allan Hills collection as a proxy for Antarctic meteorites as a whole. From our ongoing survey of Antarctic meteorites we have observed that the relative abundance of rapidly cooled H chondrites varies from site to site. We have noted that the rapidly cooled H chondrites are common at the Allan Hills sites and at the Lewis Cliff Upper Ice Tongue, rare at the Yamato icefield, and virtually absent at the Lower Ice Tongue and Meteorite Moraine at Lewis Cliff [9] and all icefields in the Elephant Moraine region [13]. We have previously suggested that this variation in abundance reflects the average terrestrial age of the populations of H chondrites from each icefield, although the exact ranking of the icefields is somewhat arbitrary because (a) the cosmogenic radionuclide database for most of these fields (at least for ordinary chondrites) is very small or nonexistent and (b) even where such data are available, they typically provide only upper limits. Nonetheless, it appears that the average terrestrial age for meteorites from the Allan Hills is about 200,000 years [14], compared to about 50,000 years for Yamato meteorites and perhaps 20,000 years at the younger Lewis Cliff sites [14]. Therefore, it appears that the rapidly cooled H chondrite group went from contributing about half the H chondrite meteorite flux to nonrepresentation in the flux over a time span of about 180,000 years. Lipschutz [4] made a similar observation on the basis of trace element data for H chondrites from only the Allan Hills and the Yamato sites. It should be noted that we have failed to find rapidly cooled H chondrites at any of the non-Antarctic find concentrations that we have examined thus far (including Roosevelt County, finds from the western U.S., and the Sahara desert). Most meteorites from these sites have terrestrial ages of <40,000 years, and most of these sites are dominated by meteorites with terrestrial ages <10,000 years [16,17]. A better test of the apparent change in the nature of the H chondrite flux over a fairly short time period would be a direct comparison between distributions in terrestrial ages for rapidly cooled H chondrites and normal Antarctic H chondrites, both with 8 Ma exposure ages. Such a comparison is hampered by the small number of Antarctic meteorites for which there is induced TL, metallographic, cosmogenic noble gas, and radionuclide data, all of which would be required for each meteorite in this analysis. Even more crippling to such a study, however, are the limitations of the terrestrial dating techniques, since the age range of most interest, between about 40,000 to 200,000 years, is just above the upper limit of 14C. It is possible to estimate ages in this range using 36Cl [14] but the uncertainties inherent in these data in this age range or so great as to make any comparison very difficult. Nonetheless, we show a first attempt at such a comparison in Fig. 2, using the 36Cl-dominated databases of Nishiizumi [14] and some recent data of Lipschutz [18]. In this figure we subdivide the meteorites on the basis of induced TL into the normal and rapidly cooled groups and, considering that most of these meteorites are from the Allan Hills collection, we expect that most of the normal H chondrites have CRE ages >18 Ma, while the rapidly cooled meteorites probably all have CRE ages of about 8 Ma. It should be noted that, in both distributions, the apparent "peak" at about 70 ka represents the lower limit of 36Cl- derived terrestrial age estimates and that the meorites in this peak may actually be spread among lower terrestrial ages. The data are too sparse for any statistical treatment, but it appears that most rapidly cooled meteorites have terrestrial ages <100 ka and, in contrast to normal H chondrites, very few have terrestrial ages >120 ka. Our present ideas about the changes in the H chondrite flux over the last few hundred thousand years are summarized in Fig. 3. Meteoroid bodies of normal H chondrites have been part of the meteoroid flux for a considerable period of time, as evidenced by their long cosmic ray exposure ages (>18 Ma). At about 8 Ma, many meteoroid bodies of both normal and rapidly cooled types were produced. We have suggested that the meteoroids of the rapidly cooled H chondrites evolved to Earth- crossing orbits faster than the normal H chondrites from the same event due to their smaller size, but it could also be argued that this difference reflects asteroid source region differences [3]. In any case, as is apparent from the terrestrial age data (Fig. 2), at about 50,000 to 100,000 years ago the H chondrite flux was dominated by rapidly cooled H chondrites. Shortly after 40,000 years ago, however, their abundance in the meteorite flux dropped considerably and normal H chondrites, including those from the 8 Ma event, dominated the flux. By about 20,000 years ago the rapidly cooled H chondrites had ceased to make any contribution to the meteorite flux and this has continued up to the present time. In order to examine the orbital history of the H chondrite groups in more detail, we have also examined their 26Al and natural TL distributions; we discuss only the 26Al data here. Using data from Evans and Reeves [19] and the Antarctic meteorite database [20] and classifying meteorites to normal and rapidly cooled groups using only induced TL, we find the distributions shown in Fig. 4. It is apparent that the normal H chondrites have a 26Al activity distribution similar to the modern falls, with perhaps a slight excess in meteorites with low 26Al activities, which would be expected in a group of meteorites which exhibit a broad range of terrestrial ages (Fig. 2). The rapidly cooled H chondrites, however, have a distribution which is significantly different than either normal Antarctic H chondrites or modern falls, with all rapidly cooled H chondrites having high 26Al activities. This is in agreement with the generally small terrestrial age of these meteorites (Fig. 2) but the activities of many of the meteorites are higher than even the modern falls. These differences may reflect the smaller size during irradition of the rapidly cooled Antarctic meteorites [9] but could also reflect different external radiation fluxes for these meteorite populations, perhaps as a result of high inclination orbits. In closing, we would note that there are many facets of the Antarctic H chondrites which could and should be explored in more detail. More terrestrial age data would be extremely useful, although their utility is, as noted above, somewhat limited by the lack of accurate dating techniques in the age range of interest. Perhaps more interesting would be an attempt to document the rise in abundance of the normal H chondrites from the 8 Ma event (such as ALHA 81037) over the last few tens of thousands of years. Only the modern falls in this group have been studied in any detail and this group is generally not found in Antarctic studies due to the tendency for these studies to concentrate on meteorites from the Allan Hills. We would suggest that H chondrites from Elephant Moraine, Yamato, and selected portions of the Lewis Cliff site would be ideal for this study. Acknowledgements. We wish to thank the many people with whom we have had discussions about this topic, especially L. Schultz, M.E. Lipschutz, S.F. Wolf, and A. Rubin. We thank the Meteorite Working Group of NASA, M. Lindstrom, R. Score, and the National Museum of Natural History (US) for samples and V. Yang and J. Wagstaff for access to the Johnson Space Center electron microprobe. This study supported by NSAS grant NAG 9-81. References. [1] Harvey R.P. and Cassidy W.A. (1989) Meteoritics 24, 9. [2] Huss G.R. (1991) Geochim. Cosmochim. Acta 55, 105. [3] Wasson J.T. (1990) Science 249, 900. [4] Lipschutz M.E. and Samuels S.M. (1991) Geochim. Cosmochim. Acta 55, 19. [5] Cassidy W.A. and Harvey R.P. (1991) Geochim. Cosmochim. Acta 55, 99. [6] Schultz L., Weber H.W., and Begemann F. (1991) Geochim. Cosmochim. Acta 55, 59. [7] Kallemeyn G.W., Krot A.N., and Rubin A.E. (1993) Meteoritics 28, 377. [8] Benoit P.H. and Sears D.W.G. (1992) Science 255, 1685. [9] Benoit P.H. and Sears D.W.G. (1993) Icarus 101, 188. [12] Benoit P.H. and Sears D.W.G. (1993) Lunar Planet. Sci. 24, 91. [13] Benoit P.H., Roth J., Sears H., and Sears D.W.G. (1994) J. Geophys. Res. 99, 2073. [14] Nishiizumi K., Elmore D., and Kubik P.W. (1989) Earth Planet. Sci. Lett. 93, 299. [15] Fireman E.L. (1990) LPI Tech Rep. 90-03, 82. [16] Benoit P.H. and Sears D.W.G. (1994) This meeting. [17] Jull A.J.T., Donahue D.J., Cielaszky E., and Wlotzka F. (1993) Meteoritics 28, 188. [18] Michlovich et al. (1994) J. Geophys. Res., in press. [19] Evans J.C. and Reeves J.H. (1987) Earth Planet. Sci. Lett. 82, 223. [20] Score R. and Lindstrom M.M. (1990) Ant. Meteorite Newsletter 13(1). Figure 1, showing Antarctic H chondrites, appears here in the hard copy. Figure 2 appears here in the hard copy. Figure 3 appears here in the hard copy. Figure 4 appears here in the hard copy. Miura Y.* Difference of Three Meteorite Groups in Orbits and Fall Time Thursday, July 21, 1994 FUTURE PLANS 5:00 - 6:00 p.m. Chair(s): M. M. Lindstrom Delisle G.* Harvey R. P. Kojima H. Pillinger C. T. Reid A. M. Contributions to Discussion of Future Plans Thursday, July 21, 1994 PUBLIC LECTURE 7:30 p.m. Chair(s): G. Delisle L. Schultz Delisle G.* Schultz L. Meteorite aus heissen und kalten Wusten: Suche und wissenschaftliche Bedeutung Friday, July 22, 1994 FIELD TRIP THROUGH THE RIES CRATER 9:00 - 5:00 p.m. Chair(s): M. Schieber G. Posges Posges G.* Schieber M. Impact Crater Nordlinger Ries--Excursion Guide 1. Introduction Both the Ries crater and the Steinheim Basin (smaller impact structure of 3.5 km diameter and some 40 km SW of the Ries; s. Fig. 2) are located in Southern Germany in the centre of a triangle formed by the geographical positions of the cities of Munich, Stuttgart and Nuremberg (s. Fig. 1). Both craters are the only impact structures in this area, despite many other smaller "craters" have been proposed. Fig. 1: Geographical location of the Ries The scientific investigation of this circular structure started 200 years ago. During that period of time many theories have been developed to explain the origin of this "hollow" in the course of the south german jura mountains. The depression divides this area in the "Suevian Alb" in the SW and the "Frankonian Alb" in the NE (s. Fig. 2). For more than a hundred years the interpretation of the volcanic origin of the Ries (the term has its source in the roman name for the "provincia raetia") has been generally accepted. The circular shape of almost 25 km in diameter and the occurence of a rock containing melted material (suevite) were reasons enough to follow this idea. Many other concepts about the origin of the basin have been made in the past. None of them, however, has been able to explain all geological, mineralogical and geomorphological features in an acceptable combination, to get a unique idea for the explanation of the Ries origin. The concept of the impact of an extraterrestrial body, first mentioned in 1904, could finally be proofed by the investigations of Shoemaker and Chao in the early sixties by the analysis of COESITE as a high pressure polymorph of quartz in local rocks. Together with STISHOVITE these minerals are typical for rocks in impact craters. The Ries impact took place almost 15 Mio. years ago in the Miocene. Coming from space an asteroid of almost 1 km in diameter (stone meteorite) hit the earth with a velocity of 70 000 km/h (= 20 000 m/sec) at least and created an impact structure of almost 25 km in diameter in a very short time (Fig. 2). Fig. 2 : Block diagram of the Ries Crater Due to the high velocity and the adequate amount of energy (equivalent to 18 000 megatons of explosives) the asteroid exploded in a huge fireball and evaporated. Approximately 150 km^3 of rock have been destroyed by mechanical breakup, melting processes and even evaporation, caused by extreme high shorttime pressures and temperatures. All life forms inside a circle of about 50 km around the impact point have been blotted out. The broken and molten rock came to a deposit inside and around the crater to form vast blankets. The occurrence of an inner ring of lifted basement material (transient crater) is a typical feature for complex craters. The crater itself has been fill again during the geological times after the impact (lake period) by various types of sediments (clay, sands, lacustrine lime). Later the removal of a part of this sediments has been caused by tectonic movements at the end of Tertiary Period by the incision of the rivers. The shaping of the basin and its surrounding areas is the result of morphodynamic processes during the Pleistocene period. 2. Description of quarries and outcrops Please see Fig. 3 for the locations of quarries and outcrops. No. 1 : Two places for a panorama view across the crater No. 1A : Wallerstein outcrop (lacustrine lime) The Wallerstein scar is located in the centre of Wallerstein, a small city 4-5 km NW of Nordlingen. It offers a fine panorama view across the crater. For the top an elevation of 495 m about sea level (related to the level of Amsterdam) has been determined. The scar juttes out about 75 m about the present basin plane. Under clear weather conditions various morphological aspects can be observed: - the Harburg port in the SE, where the Wornitz river leaves the crater basin in a narrow valley through ejected rocks, forming the crater rim - the best developed part of the crater rim in the E and especially in the S with max. elevations of about 650 m about sea level (Nordlingen: 435 m) - the "Zeugenberge" (monadnocks) of the Ipf near Bopfingen in the E (shaped like a table mountain; contact of the suevian jura mountains) and the Heselberg in the North outside the crater. Both these mountains show the pre impact geological situation (horizontal layers, undisturbed rocks) - the lower crater rim in the NW - the incission of the Wornitz river valley into the northern crater rim, where the river penetrates the Ries plain near Oettingen - the Hahnenkamm - mountains (beginning of the franconian jura mountains) - various cone shaped hills in the Ries plain, which belong to the inner ring (see No. 4) - the megablock - zone between the inner ring and the /structural) crater rim - the undulating Ries plain with its settlements and agricultural fields The Wallerstein scar belongs to the inner ring. The uplifted basement rocks are covered by lacustrine sediments. A drillhole positioned near the church of Wallerstein brought out a thickness of these lake sediments of about 30 m on a level of about 450 m NN. This means a total thickness of these secondary limestones of about 75 meters. These sediments are rich in fossils. The watersnail Hydrobia and the ostracode Cypris are lithogenous and indicate a low degree of salinity (soda lake). Sometimes fossilized landsnailes (Cepaea), mammal bones, bird eggs and tortoise - shells have been found. These animals have been floated into the lake, where they became fossilized. At the top of the scar and on its foot stromatolites (algae mats built by blue-green-algae = cyanobacteria) are exposed. No. 1B : Siegling quarry (not visited during excursion) The quarry (out of operation) is located approximately 4 km SW of Nordlingen and W of National Road B 466 (to Ulm) at the northern slope of displaced and accumulated jurassic limestone. At the eastern side of the road Multicoloured breccia (Bunte Breccie) overlies upper jurassic allochthonous limestone. The quarry location offers an impressive view along the southern crater rim. No. 2 : Wengenhausen quarry (shocked basement, lake limestone) The quarry (out of operation) is located 7 km NW of Nordlingen close to the National Road B 25 (Nordlingen - Dinkelsbuhl). It is exposed at the contact of the megablock-zone and the inner ring uplift. The rocks of this quarry are of two different types: a) shock induced, heavy destroyed and weathered rocks of the crystalline basement (gneisses, amphibolites, granites). This block consits of ejected rock material without any contact to the lifted basement. Dyke rocks occur in form of socalled (Flecken - Kersantit). Evidence of impact also comes from the occurrence of shatter cones in this dyke rock. b) The ejected basement material is partly overlain by a thin crust of a lake limestone. The bioherm-facies (algae-limestone) is apparent with a certain amount of gastropodes (land and lake fauna). The lake fauna indicates an alkaline water quality. No. 3 : Alte Burg quarry (suevite) The old suevite quarry of Alte Burg (the term means "Old Fort"; quarry out of operation) is located 5 km SW of Nordlingen close to the crater rim at the territory of Baden-Wurttemberg. The quarry is important in two respects: a) Research history The deep weathered suevite is exposed in a vertical contact with malmian allochtonous limestone (stratified displaced megablock).In former times this contact has been explained as the exposure of a volcanic pipe filled with the "Ries Trass" (= suevite) as a volcanic rock. In fact, this quarry represents the type locality for the theory of the "volcanic" origin of the Riescrater. b) Building history Besides other buildings in the city of Nordlingen and elsewhere the St. George church has been completely built with suevite. The rock for this building has been broken in the quarry of Alte Bürg. No. 4 : Wennenberg outcrops (inner ring) The Wennenberg is located 10 km E of Nordlingen as a cone shaped hill, partly covered with forest. It belongs to the inner crystalline ring with a continuous contact to the basement (lifted rock, no ejection). In this basement rocks, exposed in a small quarry out of operation, "Wennenbergit" (a type of kersantite; s. No. 2) occurs as a dyke in porphyric biotite granite and amphibolite matrix. Besides this exposure "Wennenbergit" is only known as fragments of bedrocks in the eastern part of the crater. The crystalline basement rocks are covered by lacustrine sediments, especially exposed at the eastern side of the hill. Again the occurrence of the watersnail Hydrobia and the ostracode Cypris (the fossils are rockforming) indicates a low degree of salinity of the former crater lake. A 30-minutes walk around the hill offers another possibility for a panorame view across the crater, especially to its eastern part. No. 5 : Otting quarry (fallout suevite, Bunte breccia) Suevite quarry out of operation in the moment, located NW of the village of Otting, approximately 5 km E of Wemding and the crater rim (distance to Nordlingen: 25 km). On account of its hydraulic properties suevite is used as an additive for the production of cement. Concrete produced with this cement improves its restistance against corrosion (e.g. compare the use of pumice from Santorini Island/Greece during the construction of the Suez Channel). The Otting quarry is one of the best known and largest in the vicinity of the Ries crater to study suevite and the typical post impact stratigraphy. The rock sequence is characterized by underlying limestone (Malm delta) covered by a striated surface slide caused by the displacement of huge blocks of rock. The surface slide is overlain by Bunte breccia, followed by suevite. As revealed by some drillholes especially in the northern part of the quarry, the Suevite has a maximum thickness of 24 m, the Bunte breccia of only a few meters to almost 200 meters east of the quarry (see below). The suevite quarry of Otting is well known for the possibility to collect fresh, unweathered suevite samples, containing vesiculated glass bombs and basement rock fragments (up to 98 vol. %) of all stages of shock metamorphism (see Tables 1,2). The basement material consits of magmatic rock (granite) and of metamorphic rock (gneisses, amphibolite). The portion of sedimentary rock is only 2 % or even less with mostly very small fragments. Partially fine distributed lime of secondary origin occurs forming calcite crystals at the walls of the vesicles. Concerning research history the quarry is important with regard to two aspects: a) In 1960 Eugene Shoemaker collected some rock specimens here to delivere them to Ed Chao. By analysing these samples Chao found shocked quarz-bearing crystalline rock portions which contained high pressure polymorphs of quartz: COESITE and STISHOVITE. This result was decisive for the explanation of the Ries crater as an impact structure. b) It turned out that the vesicles in the glass bombs (molten acidic basement material) contains the noble gas 40Ar. This gas is a decay product of 40K (half life time 1 300 Mio. years). By comparison the amount of both these elements the age of the impact structure could be determined to 14.8 +/- 0.7 Mio years (Gentner et al., 1963). A more accurate age of 15.0 +/- 0.1 Mio. years has been found by Staudacher et. al (1982) by using the 40Ar-39Ar method. Underlying the suevite the ejected sediments (Bunte breccia containing brecciated rock of upper-triassic and jurassic age) occur in various thicknesses up to almost 200 meters. This is the case where the pre- impact Main river valley (today a Rhine river tributary) has been filled and dammed by the ejected sediments. This former river valley had a more or less north-south-headed course with its mouth into the Danube near the city of Donauworth. No. 6 : Polsingen quarry (impact melt breccia; not visited) Small old quarry out of operation in socalled "Red suevite" near the village of Polsingen, some 5 km NW of Wemding. The melt breccia is quite remarkable, for the whole matrix consists of molten material. This fine grained matrix ist loaded with crystalline rock fragments of various stages of shock metamorphism. Inclusions of sedimentary fragments have not been observed. Coesite occurs. The typical red colour of the vesiculated melt matrix is caused by fine distributed hematite. This rock, which once has been explained as component of a lava lake of volcanic origin, only occurs in a similar form here and near Amerbach (S of the crater), which is not more exposed. No. 7 : Aumuhle quarry (fallout suevite overlying Bunte breccia) The quarry is located at the northern crater rim 2.5 km NNE of the city of Oettingen. It is under operation (suevite is used as an additive for cement production). The suevite of this quarry overlies Bunte breccia which shows a hummocky surface structure. It contains numerous glassbombs with a sometimes reddish colour. Basement fragments are abundant. The corn size of the matrix, the sizes of the glass bombs and the basement fragments are under a wide range of variation. The rock ist partly deep weathered with a rather frequent occurrence of degassing pipes (here yellowish montmorillonitic clay minerals). The Bunte breccia consists of various sediment fragments (clay, sandstone) of upper triassic lower and mid jurassic age. The colours show a big variety: purplish-red and white for the upper triassic sandstones, dark coloured lower jurassic clay and red mid jurassic sandstone. The roll-glide-mechanism is clearly visible by turbulent features in the sediments. The rock contains a lot of fossils, especially the shocked and sliced gaines of the so-called "Ries- Belemnites". No. 8 : Buschelberg quarry (lacustrine lime) The quarry (out of operation) is located at the northern crater rim near Hainsfarth, 2 km E of Oettingen. Quarry under nature conservation. The wide range quarry represents one of the largest distributions of exposed lacustrine lime of the Ries-crater. On surface the quarry walls are 6 to 7 m high. Two different types of facies are exposed as this northern margin of the crater lake: - an algal bioherm facies, represented by root shaped bodies which can be combined to greater cone shaped components (bundles) of sometimes 1 to 2 m in diameter - between the bundles a sedimentary facies is developed, formed by the accumulation (rock forming) of numerous individuals of the watersnail Hydrobia and the Ostracode Cypris (again an indication for a low degree of water salinity). For further informations about the geological structure of the crater, the distribution of rocks, the description of rock types, their characteristics and the stages of landscape development we refer to the list of references and to Figures 4,5 and 6a-d and Tables 1,2 and 3. Fig. 4 Fig. 5 Fig. 6 a - e Table 1 Table 2 Table 3 3. References The titles cited here are only a small, but representative collection of publications about the Ries impact crater, the Rieskrater Museum Nordlingen, field guides and maps. - Bayerisches Geologisches Landesamt (ed., 1977a): Ergebnisse der Ries- Forschungsbohrung 1973: Struktur des Kraters und Entwicklung des Kratersees; Geologica Bavarica 75, 470 pp, Munchen (here extensive list of references) - Bayerisches Geologisches Landesamt (ed., 1877b): Erlauterungen zur Geologischen Karte des Rieses 1:50 000; Munchen (incl. geological map) - Chao, E.T.C. et al. (1987): Aufschlusse im Ries-Meteoriten-Krater; 84 pp; Bayerisches Geologisches Landesamt, Munchen (english printed version available, incl. geological map 1:100 000) - Fischer, K. (1980): Das Nordlinger Ries und seine Nachbarlandschaften ; in: Rieser Kulturtage Dokumentation, Vol. III, 365 - 379 - Fischer, K. (1990): Flussgeschichte und Reliefgenese im Norden und Osten des Nordlinger Rieses; in: Rieser Kulturtage Dokumentation, Vol. VIII, 60 - 82 - Gentner, W. et al. (1963): Die Kalium-Argon-Alter der Glaser des Nordlinger Rieses und der bohmisch-mahrischen Tektite; in: Geochim. Cosmochim. Acta 27, 191 - 200 ,- Graup, G. (1990): Impact Craters of Nordlinger Ries and Steinheim Basin; Guide for Excursion 3B, International Volcanological Congress, Mainz; 52 pp; unpublished. - Huttner, R. (1990): Zur Geologie des Rieses (Exkursion G am 20. April 1990); in: Jahresbericht und Mitteilungen des oberrheinischen geologischen Vereins, Neue Folge 72, 157 - 175 - Kavasch, J. (1991): Meteoritenkrater Ries - ein geologischer Fuhrer, pp 112 (engl. printed edition available) - Pohl, J. et al. (1977): The Ries impact crater; in: Roddy, D.J. a.o., Impact and explosion cratering, 343 - 404, New York (here extensive list of references) - Posges, G. and M. Schieber (1994): Fuhrer durch das Rieskrater - Museum Nordlingen; in: Das Ries, Akademiebericht Nr. 253 der Bayer. Akademie fur Lehrerfortbildung Dillingen, 1 - 91 (here extensive list of references). - Posges, G. and M. Schieber (1991): Das Rieskrater - Museum Nordlingen und sein geologisches Umfeld; in: Archaeopteryx 9, 83 - 87 - Schieber, M. (1989): Soil formation in displaced pleistocene aeolian sands; in: Catena Supplement 15, 269 - 278 - Staudacher,Th. et al. (1982): 40Ar - 39Ar ages of rocks and glasses from the Nordlinger Ries crater and the temperature history of impact breccias; J. Geophys. 51, 1 - 11 - Stoffler, D. (1972): Deformation and transformation of rock forming minerals by natural and experimental shock processes; in: Fortschritte der Mineralogie, 49, 50 - 113 - Stoffler, D. (1974): Cratering mechanics, impact metamorphism and distribution of ejected masses of the Ries structure - An introduction; Fortschritte der Mineralogie, 52, 109 - 117 - v. Engelhardt, W. (1974): Ries Meteorite Crater, Germany; Excursion Guide B4; in: Fortschr. Min., 52, Beiheft 1, 103 - 122; Stuttgart TITLE ONLY Marvin U. B.* A Historical Outline of Meteorite Discoveries in Australia and Antarctica Australia with its tectonic stability and large areas of arid to semi- arid climate with low population density preserves the most remarkable record of impact features, meteorites, and tektites of any area of similar size in the world. Antarctica with its vast, shoreward-creeping ice sheet has yielded the worlds most extensive collection of meteorite fragments. These two southern continents exemplify the importance of hot and cold deserts as places to search for evidence of collisions between the Earth and bodies from space. Nevertheless, in both continents the collection and study of meteorites began much later than elsewhere. This paper will briefly outline the history of meteorite and related discoveries in Australia and Antarctica. Australia Tektites. Archaeological studies show that the aboriginal peoples used australites in their rituals from prehistoric times. The aborigines told the early settlers that these glassy bodies had fallen from the sky. The first description of an australite was written in 1844 by Charles Darwin [1], who had been sent a beautifully flanged specimen collected on a sandy plain between the Darling and Murray Rivers in New South Wales. From its "singular artificial-like" appearance Darwin concluded that it was a volcanic bomb that had burst open and spun rapidly in midair. Knowing that it had been found hundreds of miles from any volcanic region, Darwin commented that it must have been transported either by the aborigines or by natural means. In 1892, Victor Streich sent australites he had collected on the Lindsay expedition through the western deserts to A. W. Stelzner in Germany. Streich thought the tektites were a type of meteorite, but in 1893 Stelzner described them as volcanic in the belief that meteorites contain no glass. Four years later, R. D. M. Verbeek, who had calculated the force of the Krakatau eruption of 1883, proposed that australites and all other tektites known at that time were ejected by volcanoes on the Moon. This idea gained widespread support in Australia and Europe [2]. In the early 1960s, after the opening of the Space Age, interest in tektites soared and new types of chemical and isotopic analyses ultimately led to the current consensus that these bodies are splash glasses from hypervelocity impacts on the Earths surface. Few investigators dispute this mode of origin, although details of the process remain poorly understood and the elegantly flanged australites, which show clear evidence of two stages of melting, present special problems. In any case, we should note that the Australian aborigines were correct in their belief that the tektites they so prized fell out of the sky. Meteorites. Unlike tektites, iron meteorites apparently were not put to use by aboriginal peoples anywhere in Australia [3]. The first well- documented discovery of meteorites in Australia occurred as late as 1860 when a supposed deposit of iron found in 1854 near Cranbourne, Victoria, proved to consist of two large, mostly buried, iron meteorites weighing 3.5 and 1.5 tons. (The Baratta stony meteorite may have been discovered in 1845, but documentation is lacking). Samples of the Cranbourne irons were sent to W. K. von Haidinger in Vienna who published the first descriptions and analyses of an Australian meteorite in 1861. Twenty- four more meteorites, of which 16 were irons and one was a pallasite, were found in Australia during the remaining years of the 19th century. Thirteen meteorite falls, of which only one is an iron, have occurred in Australia since 1879 (Table 1). From 1860 to 1955, meteorite discoveries occurred in Australia at an average rate of about one per year [3]. A striking increase in discoveries, particularly of stones, followed the dawn of the Space Age in 1957. At first, the increase was due largely to a growth in interest and recognition skills of ranchers, prospectors, and rabbit hunters who work in the outback. More important in recent decades have been systematic searches of desert areas by teams from the Western Australian School of Mines at Kalgoorlie and the Western Australian Museum at Perth. Since 1985 the rate of discovery has risen to nearly nine meteorites per year. By 1992 the total number of catalogued Australian meteorites was 278, and counting. The richest area in Australia, and one of the richest in the world, is the Nullarbor Plain with its semi-arid climate and flat, treeless expanses of light-colored limestone where meteorites are well preserved and easily recognized. In 1991, the Nullarbor Plain yielded the first lunar meteorite to be found outside Antarctica. Impact Craters. The first meteorite craters to be recognized in Australia were a group of rimmed depressions near Henbury Station, which lay in the central part of the continent 117 km by camelback from the nearest railroad. Early in 1931 two residents of the Henbury area independently brought these craters to the attention of Professor Kerr Grant at the University of Adelaide. Their interest was sparked by widespread publicity given the fall of the Karoonda meteorite from a spectacular fireball in South Australia on November 25th, 1930. Sir Douglas Mawson, then serving as honorary mineralogist at the South Australia Museum, recommended that the reports be investigated. Shortly thereafter, Arthur R. Alderman, of the University of Adelaide, visited the Henbury site and located 12 shallow craters (1 more was found later) in a semi-arid plain strewn with iron meteorites. Masses of black glass lay just outside the two largest craters. The Henbury craters proved to be crucial in gaining acceptance for the highly controversial idea of meteorite impact as a crater-forming process. Acrimonious debate still persisted on the impact vs volcanic origin of "Meteor Crater," 1.3 km across, near Winslow in northern Arizona. The only other feature for which an impact origin had been proposed was the crater, 161 meters in diameter, at Odessa, Texas, first described in 1928. Specimens of Henbury meteorites and glass were sent to Leonard J. Spencer at the British Museum, who compared them with those from the craters at Wabar, Arabia, discovered in 1932. In 1933, Spencer [4] published a landmark paper, Meteorite Craters as Topographic Features of the Earths Surface, in which he established the authenticity of impact craters as geological phenomena. Spencer argued that, far from being simple dents or pits in the ground, true impact craters result from the explosive release of energy when hypervelocity meteorites strike the Earth. By 1987 nineteen impact structures had been identified in Australia (Table 2) and searches continue for additional ones [5]. These structures range in size and age from the Dalgaranga Crater, 24 meters in diameter and less than 3,000 years old, to the 600 million year-old Lake Acraman structure 35 km in diameter. Lake Acraman is one of the most remarkable impact sites in the world. The event projected shocked fragments of a 1,600 million year-old volcanic complex for distances of 300 to 450 km where the breccia serves as a 600 million year-old horizon marker in late Precambrian sections [6]. Antarctica Early Meteorite Discoveries. The first meteorite to be discovered in Antarctica was a 1-kg stone lying on compact sastrugi in Adelie Land. It was collected in 1912 by a member of the Australasian Antarctic expedition of 1911-1914 led by the Australian geologist, Douglas Mawson [7]. Named "Adelie Land" and classified in recent years as an L5 chondrite, the main mass now resides in the South Australia Museum at Adelaide. Half a century passed before any more meteorites were found in Antarctica. In 1961 members of the 6th Soviet Antarctic Expedition, mapping a gneiss-diorite formation at an altitude of about 3,000 meters in the Humboldt Mountains, collected two fragments, weighing 8 and 2 kg, of an iron meteorite. The pieces lay about 7 cm apart on a rock- strewn surface approximately 40 meters from the edge of the ice sheet. Both were severely weathered with ridged surfaces displaying Widmanstätten structure. The larger mass also was marked by deep fissures that appeared to have been widened by a long succession of freezing and thawing episodes. This process may well have caused the smaller fragment to spall off the larger. The meteorite was named "Lazarev" for one of the base camps of the Soviet expedition and pieces eventually were deposited at the Academy of Sciences in Moscow and the municipal museum in Leningrad. In 1975, Vagn F. Buchwald [8a] wrote that certain anomalous aspects of the metal indicate that the Lazarev meteorite probably is not an iron but a metallic remnant of a pallasite from which the olivine crystals have been lost. Two pieces of an authentic pallasite, weighing 23 and 9 kg, were found lying about 90 meters apart on the surface of a glacier by a U. S. team working in the Thiel Mountains in 1962. The main mass of the "Thiel Mountains" pallasite resides in the U. S. National Museum in Washington. In 1964, two engineers in a U. S. field party found an iron meteorite, weighing just over 1 kg, among glacial cobbles about 30 meters above the surface of the ice surrounding a nunatak in the Neptune Mountains. The "Neptune Mountains" meteorite retains patches of fusion crust and appears so well preserved that it seems unlikely to have been transported very far by the ice [8b]. Perhaps it fell where it lay. Each of these four meteorites was discovered by parties pursuing other interests (Table 3). No scientist predicted that meteorites would be found in Antarctica other than by chance discoveries. Meteorites on Stranding Surfaces. The first indication that the moving ice sheet may concentrate meteorites in what amount to "placer deposits" followed from a report presented by Shima and Shima [9] to The Meteoritical Society in 1973. Their analyses showed that nine meteorites collected in 1969 by Japanese glacial geologists from a small (9 x 5 km) area of the Yamato Mts. icefields were not shower fragments but samples of four different stony meteorites. This unprecedented discovery gave rise to the concept that, under special circumstances, meteorites from diverse falls may be frozen into the ice and exposed on stranding surfaces--expanses of ice temporarily trapped behind mountain barriers and worn down by wind ablation. During Shima's presentation, Prof. William A. Cassidy of the University of Pittsburgh, decided to submit a proposal to the National Science Foundation, which funds all Antarctic research by U. S. scientists, to search for meteorite concentrations within helicopter range of McMurdo Station, across the continent from the Yamato Mts. On returning home he reported the Shimas' results to Dr. Takesi Nagata, Director of Japan's National Institute of Polar Research, who was visiting the University of Pittsburgh at that time. Nagata sent a field party back to the Yamato Mts. icefields in December, 1973, to conduct the first search specifically aimed at collecting meteorites in Antarctica. The members recovered 12 more specimens that season. It was by no means so straightforward a task to field an American team. A lapse of up to 18 months generally is expected between a scientist's submission of a new proposal to the NSF and his or her arrival in Antarctica. However, as a radical departure from traditional Antarctic programs, Cassidy's proposal received mixed reviews and was declined. Many of the geologists, glaciologists, and other scientists who dominated Antarctic research saw meteorites as of minimal scientific value, and referees with personal knowledge of the frigid Antarctic wastes viewed the idea of focused searches for them as naive. Cassidy submitted a revised proposal in 1975 which also was declined, but he resubmitted it and it was accepted on short notice after he called the NSF with the news that a Japanese team had just returned with 663 specimens. In 1976, Cassidy led the first U.S. meteorite search in Antarctica. That year Nagata also sent a scientist to search for meteorites out of McMurdo Station and so joint searches were agreed upon with equal sharing of all specimens found. This arrangement continued for three years. With the exception of 1990 when the season's work was cancelled due to lack of logistical support, Cassidy led the U. S. teams every year until 1992 when he passed along the leadership role to his Co-investigator, Dr. Ralph Harvey. By 1994, twenty-one years after the Shimas surprised meteoriticists with their report, more than 15,000 meteorite fragments, possibly representing about 1,500 individual meteorites, have been collected from the cold Antarctic deserts by parties from Japan, the U.S.A., and Europe (Table 3). Research samples have been made available to laboratories worldwide. The inventories include 12 lunar meteorites ejected by meteorite impacts on the surface of the Moon, 4 meteorites showing persuasive evidence of an origin on Mars, and several new species of asteroidal materials. Antarctic meteorite programs have given new inpetus to planetary research, and earth scientists, some of whom initially questioned their value, have engaged in collaborative efforts to relate meteorite concentrations to a wide range of problems including the dynamics of ice motion, the configuration of bedrock beneath the ice, the ages of ice patches within meteorite stranding-surfaces, the sources of volcanic dust bands, and indications of climate changes. References [1] Darwin C. (1844) Coral Reefs, Volcanic Islands, South American Geology. (Reprint, 1910, Ward Locke & Co, Ltd. London, p. 191. [2] O'Keefe, J.A. Tektites and their Origin. Elsevier, N.Y., pp. 3-5. [3] Bevan, A.W.R. (1992) Records Aust. Museum, Supplem. 15:1-27. [4] Spencer L. J. (1933) Geograph. Jour. 81:227-248. [5] Shoemaker, E.M. and Shoemaker, C.S. (1988) LPSC XIX:1079-1080. [6] Gostin, V.A., Halnes, P. W., Jenkins, R.J.F., Compston, W. and Williams I.S. (1986) Science 233:198-200. [7] Mawson, D. (1915) Home of the Blizzard, Hodder and Stoughton, Ltd., London, 2:11. [8a] Buchwald V. F. (1975) Iron Meteorites Univ. California Press, Berkeley, 2:761. [8b] Buchwald, V. F. (1975) 3:890. [9] Shima M. and Shima M. (1973) [Abs.] Meteoritics 8:439- 440. Table 1. Australian Meteorite Falls [3] Year Locality Class 1879 Tenham Queensland L6 1895 Rockhampton Queensland Stone 1900 Emmaville New South Wales Eu 1902 Mount Browne New South Wales H6 1928 Narellan New South Wales L6 1930 Moorleah Tasmania Iron 1930 Karoonda South Australia C5 1942 Forest Vale New South Wales H4 1960 Woolgarong Western Australia L6 1960 Millibillie Western Australia Eu 1967 Wiluna Western Australia H5 1969 Murchison Victoria CM2 1984 Binningup Western Australia H5 Table 2. Australian Impact Structures, 1987 [5] Year Identified 1932 Henbury Craters Northern Territory 1933 Mt. Darwin Tasmania (confirmed 1973) 1937 Boxhole Northern Territory 1938 Dalgaranga Western Australia (noted 1923) 1947 Wolfe Creek Western Australia 1967 Gosses Bluff Northern Territory 1971 Liverpool Northern Territory 1970 Strangways Northern Territory 1973 Kelly West Northern Territory 1976 Veevers Western Australia 1980 Goat Paddock Western Australia 1980 Teague Ring Western Australia 1982 Lawn Hill Queensland 1984 Fiery Creek Dome Queensland 1986 Lake Acraman South Australia 1987 Connaly Basin Western Australia 1987 Mt. Toondina South Australia 1987 Piccanniny Western Australia 1987 Spider Western Australia Table 3. Antarctic Meteorite Finds Chance Discoveries: 1912-1969 1912 Adelie Land 1 Stone Australasian Expedition 1961 Lazarev 1 Iron or pallasite USSR 1962 Thiel Mts. 1 Pallasite USA 1964 Neptune Mts. 1 Iron USA 1969 Yamato Icefields 9 Stones Japan Directed Searches: 1973-1993 Japan 8,424 USA 6,391 Euromet 285 N. Zealand 16 Total Fragments 15,116 Total Meteorites ~1,500 (?)