 Welcome back, everybody. I think we had a really great day yesterday. And I'd like to thank all of the members. Yesterday we talked a lot about South America and Mesoamerica. And today we're moving on. We're going to be talking about Asia and Europe. And in the afternoon, we're going to talk more about analytical methods and hear a number of talks. And of course, we'll end the day with our poster session and continue that. So yeah, why don't we just kick it off right now? Our first paper in the Asia and Europe Obsidian Sourcing and Archeology session is Exploitation of Obsidian at Saituaj Rock Shelter, North Central Caucasus, Russia by Katerina Doronacheva, Lyubov Golovanova, Vladimir Doronachev, Galina Poplovko, Andrei Netomolkin, and Stephen M. Shackly. So Katerina, why don't you start it off? Good evening. Can you see my screen? Not yet. Is it you, though? One second, please. Does it work now? Yes. Yes. Good? OK. I don't have a stable internet connection. So maybe some problems with internet. Because I am in the Caucasus right now. Dear colleagues, I would like to represent our preliminary results of obsidian exploitation and representing a general news site that can make the Pleistocene holocene in the North Central Caucasus. This area is currently in the South of Etienne and called Psytoazia Rock Shelter. Like the epipallelic industries in South West Asia, the epipallelic industry of Caucasus is notable for the first appearance of the rebel geometric mechanism, which represent a distinctive tool group in the epipallelic assemblies in the Caucasus, dated from the late Glacial to the early Holocene between approximately 20 to 10,000 years. Modern archaeological data from the Caucasus also show distinct regional similarities and a specific pathway of epipallelic development in the Caucasus. The first region divided the greater Caucasus margins into the western and eastern parts also is the location of the Zayukov-Baksal obsidian soaps. Our search show that Zayukov obsidian was actively used in the paleology of the northern Caucasus, as you can see on the ground as a little bit. Psytoazia Rock Shelter, discovered by our team in 2018, is located about 70 kilometers northeast of the height European volcanic mountain peak of Elbrus and about eight kilometers south from the town of Zayukov in the Baksan River valley. The rock shelter is situated in a deep, up to 200 meters, terraced and forested valley of the Fanduka-Oseraciuco river and about two to three kilometers upstream from the middle paleologic site at Seraciuco grove. Psytoazia Rock Shelter is formed in ruralic incubates and top of the lower chig information. The rock shelter is more than 100 square meters. Psytoazia Rock Shelter likely represented in the Pasta Grotta, which was partially destroyed by late erosion. In 2018, a small test excavation was undertaken in the rock shelter. To examine the strategically of the deposits and in 2019, the excavation started at the site. They will be continued in the current season. The profile mainly represents the upper part of the rock shelter deposits. Free radiocarbon dates on bone from layer two at Psytoazia were obtained in the radiocarbon laboratory in St. Petersburg, Russia. You can see them in the slide. The date and result indicates that the final epipalytic assemblage recovered from layer two at Psytoazia Rock Shelter dated from the very beginning of the Holocene. Fallen to the Pleistocene Holocene boundary with a layer two assemblage occupies a specific position among epipalytic sites in the northern Caucasus. Fallen analysis conducted for Psytoazia Rock Shelter indicates the layer two formed under warm and dry climatic conditions. At the stage, coniferous-dominated wood associations occupied higher elevations in the area, raised homing and elm forests with participation and as treated in the Fanduko River Valley. Due to more than 1,000 bones, we found in the rock shelter, mainly represented by Caucasian wood, wild horse, and some other. Almost all of the animal species identified belong to the widespread Caucasian species that are present in the modern Mesophilic corner. The analysis of microstratigraphy and space distribution of liturgical artifacts in layer two shows that the main concentration of artifacts coincides with the distribution of bones and is confined to the uppermost 20 centimeters of layer two. The main artifact concentration is in water C10 and also water CLN. The assemblage from layer two in Psytoazia Rock Shelter comprises in total 634 artifacts made of flint and obsidian, including one tool from home. The assemblage allows us to identify two contexts that characterize the main activities of the epipalladic humans in the cell. The technological context related to the stone napkin and production of blanks for tools and the tool inventor related to subsistence. As you can see on the graph, flint was the main raw material at layer two of Psytoazia Rock Shelter. Results of petroleum and geological analysis indicate local origin of flint, probably from sources located less than 10 kilometers from Baxan River Valley and Chigembina Valley from the site. Results of the X-ray analysis indicate that the obsidian artifacts are made exclusively of obsidian originated from the Zyneco source, like each six to seven kilometers to the west from Psytoazia Rock Shelter. Only nine artifacts are made from other types of raw materials that include horns, lathe, undisputed, incubate, and sandstone. The technological context of the litic assembly from layer two includes cores, remnants, chips, core treatment elements, laminar blanks, and laminar and original flakes. Among the 11 cores, unipolar cores prevail. The 41 technical plates or core treatment elements indicate active onsite production and reduction of cores. The layers to assemblage also include full and flint artifacts that are related to stone snapping. There are two, the vast majority of laminar blanks are fragmented. Complete laminar blanks are not humorous. Their length range from 12 to 65 millimeters and blanks longer than 40 millimeters are rare. In general, the analysis of natural characteristics of laminar blanks suggests that the mechanical technology and the reaction of the light is increased from five to 12 millimeters. The assemblage for layers to lips to the shoulder comprises a total of 56 tools, including the two tools described above. Points include fragments of wash-on points and the shoulder end point with the shoulder formed by a branch attached. But the lights are mainly represented by fragments. Abundant truncated tools are made by abrupt retouching blades and bloodlets. Geometric microlays are quite humorous but include only tangles. Most endscrappers are made on blades. One endscrapper is made on a crested bloodlet and two endscrappers are made on flakes. A single variant found in layer two is made on a bloodlet and can be defined as a double-angle term. You also report the first preliminary results of use wear analysis of print and obsidian tools from layer two at Psittaja-Rickshaw. The research is ongoing due to the hard illness of our use wear specialists. The goal of use wear studies is a detection of a possible use of artifacts and photography of use wear macro traces. All artifacts have a very good preservation. A small sample of 15 obsidian and print artifacts selected from layer two assemblage at Psittaja was used for the preliminary physiological study. The method developed by Semenov was used. The study allows distinguishing different categories of tools that have been used as hunting weapon and for battering hunting breed brought by the Epipallelitic inhabitants to Psittaja-Rickshaw. Traces of use wear were found on 10 tools analyzed for the management. Seven from print and three from obsidian. On which traces of use wear of the common comparations are found have a good and have good preservation. We defined mid knives with traces of use as mid knives on four tools. The age of work and zones of the knives is stronger, smooth, has a characteristic armor shape and there are characteristics similar to fractures. Scrapers on bone and leg, five tools we defined and scrapers on bone and leg. And finally one obsidian tool had the worst preservation. But this was a point probably exploited as a spearhead and secondary use as percor on skin. Also on some artifacts there were identified macrotraces of wear. Here you can see a table that demonstrates different types of raw materials and categories of tools. Also on some artifacts there were identified macrotraces of wear, smoothen and abrasion, resulted from the two mountain in a wooden shop. Analyzing my report, the technical anthropological characteristics of the medical sandwich from layer two with a terrible shelter described above are consistent with the path of epilogic age defined earlier for this assemblage. We then on the assemblage composition and the tool set represented in layer two and taken into account the remote location of the site from other epilogic sites known in the area. The Psythagoric Shelter Cave site can be considered preliminary as a human habitation campsite that was probably used as a hunting camp. Like other epipolagic sites in the Caucasus, the epipolagic humans at Psythagoric Shelter subsisted mainly on Caucasian war while the couple had set great seasonally which the epipolagic hunters of the Caucasus dug at elevations below 1500 meters above the sea level both in the forested and open environments. Also, they were found tools used for butchering hunting prey, meat knives, bone antler work. Give this character some bone antler and probably hunting. For preliminary results of analysis of liturgical and phonological assemblages indicated in layer two humans were engaged in intensive naping of print and obsidian and production and use of tools made mostly from these two types of raw materials for butchering and consumption of hunting prey presented by angled animals. Thank you. Thank you, Katarina. Does anyone have any questions? No questions? No questions. I have a question. Are you exploring protein residue on the knives, animal protein residue? Yes, we work in the direction with both middlapalagic sites. These are the same very close to each other. And this side, yes. But as a result, I think. Great. Kate, another point to make here is that the fact that all of the obsidian is from Zyacobo is relatively unusual for some of these sites at that age. I mean, other sites have a mix of sources. In the Zayukova-Baksan source area, we tested for the moment three ethypallelitic sites that covers almost all ethypallelitic sites in the area. And in all the sites, only Zayukova is here. But in this mask, I think, I'm located directly to 150 kilometers from this modern relief. So it's about 650 kilometers to the modern relief. We have both Zayukova and the Chikiani obsidian in the apopanelitic labs. Yeah, we were talking about obsidian that has mahogany color, red color earlier. Nico and I were talking about it. And Zayukova has a lot of red obsidian. So I think that probably increased its value. Yeah, very beautiful things from red obsidian because it's the middle ethypalelitic assembly of Zayukova. All right. Well, the next talk is also by Katerina. And I'll just ask, we really heard you best when you spoke directly into the microphone. So for this, I think the sound is much better when you speak more directly into the microphone. I will try it, but I have only my microphone in the computer. I don't have it in my hands. Certainly. All right. Well, let's kick it off then. We have obsidian exploitation in upper paleolithic, layer 1A-2 at Mesmeskaya Cave, northwestern Caucasus, Russia, by Katerina Doronacheva, Steven M. Shackley, Leobov-Golvanova, Vladimir Doronachev, and Galina Pulplovko. Take it away. Thank you. Currently, there is much attention being paid to the study of raw material exploitation during the polarity. Studies of raw material use, our ability of raw materials in archeological collections, and the identification of raw material sources provide new data on lithic technology, territoriality, and adaptations from the ability of social and free papers. However, until recently, raw material exploitation has been never studied specially for the paleolithic sites in the Caucasus. In my presentation, I would like to introduce you shortly to the main results of the recent studies of obsidian exploitation in the upper paleolithic layer 1A-2 at Mesmeskaya Cave. Mesmeskaya Cave is located 1,300 meters above sea level in a small tributary of the Kuzhips River. This is Kuban River Basin, about 50 kilometers south of the city of Maipok in the northwestern Caucasus. It is more than 500 square meters up to 10 meters in height and faces southwest. So this is a huge cave. Since 1987, when the local team started excavations on the site, over 100 square meters have been excavated to the maximum depths of 5 meters. The lowest Pleistocene layers 4 to 7 are excavated only in a test pit, contained no archeological material. The Pleistocene strata are most completely preserved towards the interior of the cave. Near the cave entrance, heterogeneous erosive processes have destroyed upper paleolithic strata, and the Holocene layers, unconfirmable, underlies the middle paleolithic deposits. Since 1997, eight upper paleolithic and api-paleolithic layers have been identified in the excavations in line 1829 in the interior part of the cave. From top to bottom, 1-3, 1-4, this is api-paleolithic. And there's 1-1, 1-1 slash 1-2, 1-2, 1-b-1, 1-b-2, and 1-c, these are upper paleolithic, including only upper paleolithic layers. The new switch is found strata 1-d. Line atrop the upper most middle paleolithic layers down, and contain no archeological funds. The only insidious api-paleolithic layer 1-a, sorry, 1-a frame, as the lower api-paleolithic layer 1-4 is disturbed and eroded because the niche is not represented in more recent excavations deeper in the cave. All upper paleolithic layers are inclined to the north-west, towards the left wall of the cave, and preserved in the cave. The early api-paleolithic assemblage from Musmaeskaya cave layers 1-c to 1-a, dated by a robust series of radiocarbon and estimates between 33 to 27,000 years, are characterized by a developed bladeless technology and a range of tools made on bladeless and microbladeless, including numerous microbladeless and microbladeless points. Blades, bladeless, and microbladeless predominate among total flakes. Bladeless and microbladeless are most common on laminar blanks, while blanks are relatively infrequent. These industries are similar to the South Caucasian early api-paleolithic and Levantinochmarin. Bond tools and ornaments are numerous and well-developed. At Musmaeskaya, 25 radiometric, mostly AMS, dates have now been obtained by five different laboratories for the api-paleolithic layers, including more recent AMS determinations using ultra-filtration obtained by the Oxford Radiological Accelerator Union from humanly-modified the Akhat-Mang Falls. One previous conventional date defines the calendric age of layer 1 and 2, approximately 27,000 years, obviously in the same chronological range as the younger estimate for layer 1B. So far, more than 42,000 litric artifacts, including material derived from water sieving, have been found in all api-paleolithic layers at Muska Cave, as you can see on the table. The bigger collection comes from layer 1 and 2. More than 60 obsidian artifacts in total have been analyzed using the XRF method from all api-paleolithic layers at Musmaeskaya Cave. Among them, 32 samples from layer 1 and 2. The analysis were conducted on a thermo-scientific quantity in the XRF spectrometer. The spectrometer is currently located in the gearcheological XRF fluorescence laboratory in Albuquerque. The sources studies identified that only two obsidian sources, the Yukovo Obaxan, located in the north-central Caucasus, about 80 kilometers north-east from Mount Elbrus and about 250 south-east from Musmaeskaya, and Chiki, located on the Djavaheti plateau in southern Georgia, in the Lasso Caucasus, about 450 kilometers south-east from Musmaeskaya, were used exclusively by the api-paleolithic humans. Our study shows that at Musmaeskaya, the api-paleolithic humans in layer 1 and 2 have tolerated more intensively the Chikiani-Paravani obsidian source area, rather than the flow of the Yukovo source in the north-central Caucasus, contrary to the low-middle-paleolithic neanderthals for which our study identifies only exploitation of the Yukovo Obaxan source. Most of the obsidian artifacts at Musmaeskaya come from layer 1 and 2. In this collection, they comprise more than 2 percent in total sandwich. The overall composition of the obsidian artifact sandwich, which includes one core, are co-treatment elements, cortical flakes, flakes, and numerous fragments and chips, such as the telemetry downside napion of obsidian loop plate, laminar blanks, are dominated by bloodlets and micro-bloodlets. The composition of detached tools from obsidian is characterized by predominance of packed and detached bloodlets and micro-bloodlets. Also, a gravid point and burren made in obsidian blanks were found. The core analyzed by XRF was made from the Yukovo Obaxan obsidian. The data suggests that in layer 1 and 2, obsidians were apparently transported to the cave not only in the form of ready-to-use blanks, bloodlets or micro-bloodlets, or detached tools, that have been subsequently retouched inside or re-driven in the cave. While also, there were pieces of raw material for subsequent changes in obsidian transportation and use in layer 1 and 2, compared to previous layers. Also, we think more data are required for more confident comparison and conclusions. The onsite obsidian tool production in layer 1 and 2 is additionally confirmed by the presence of obsidian artifacts concentrated on quadrant M11. You can see this on the slide. It includes 63 obsidian artifacts. Also, no cores were found on the square. There are three quadrant elements, five flakes, including cortical ones, two flake fragments, 26 microchips and eight micro-bloodlets fragments, and some other fragments and one tool burren all produced from obsidian. Apparently, this concentration suggests by products in the area where an obsidian core was flake to produce a series of legs. Probably micro-bloodlets subsequently were detached into tools and transported away. The upper polarity layers have been excavated over different areas due to the lower upper polarity layer 1c and 1b. Apparently, it appears already in layers 18 and 17. It was excavated since 1997 while the trope of polarity layers appears only in layers 30 and 15. It was excavated only in excavations deep in the cave. The distributions of obsidian artifacts are also various. In layer 1 and 2, it was excavated over an area of more than 40 square meters. This yielded the largest number of lithical artifacts, indicating the most intensive human occupation during the entire upper polarity can be mass-cooking. The layer also produced the largest number of obsidian artifacts. And most of them have been found mostly on lines 10 to 12 and 14. You can see this on the table. Our study suggests that the early upper polarity humans transported obsidian mostly as ready-to-use blends. Presumably, they transported chicken, obsidian as ready-to-use blends and tools or attached tools, made on bloodlets and micro-bloodlets. And as prepared course, this is for the eucalyptus. It also indicates that the early upper polarity humans more intensively exploited the Chicania Paravani source area in the southern Caucasus, rather than the closer the eucalyptus source in the north-central Caucasus. And the exploitation of the South Caucasian source increased towards the end of also our studies of obsidian and flint artifacts, from in my case, suggest a deliberate selection of high-quality raw materials derived from distant sources by the early upper polarity humans. In the cave, local low-quality flints were used mainly for making flakes and only relevant for production of bloodlets and micro-bloodlets. Bloodlet tools or attached tools, such as inscritor syndrome. These places of artifacts are made mainly on high-quality, non-local flints and obsidian. Thank you. Thank you, Catarina. Do we have any questions? No questions? I just wanted to say those are some impressive caves. That's some impressive places to work. Very, very nice. Yeah, for a North Americanist to analyze obsidian that's 35,000 years old, it's quite an experience. Yeah, and what's incredible when you have Neanderthal use and everything, it's really interesting. Yeah, it's pretty neat. Well, wonderful. So we're a little bit ahead of schedule. So why don't we kind of give some time to catch up? Our next scheduled talk is at 9.40. So maybe we'll take whatever, 15 minutes here and wait for the start and as we move on to Japan. So we'll take 15 minutes here. Catarina, I have one question for you. A colleague of mine, my old advisor is working on some cave sites in southern Ethiopia and they have obsidian and they're doing some hydration work on it. Have you thought about doing hydration work on the obsidian just to get a sense of how to calibrate the carbon dates you have with obsidian hydration at all? Is that something that's possible? I was just talking to Chris about it. I'm trying to reel him in. Because I have a lot of the artifacts here. Yeah, you're starting to work out. It's starting to work. You're starting to, the lines starting to tighten. Yeah, the glass is very transparent and very kind of amenable to new methodologies that are coming online. And it would be a real challenge to date something that's 30,000 years ago. And I'm not saying I can do it, but I'll give it a try. Well, it's a good experiment, I think. If you wouldn't mind keeping me in the loop so I could maybe bring some folks together. Ben Smith, who is going to give a poster tomorrow, is also, might also be curious about that work. I've been working on those materials for the past four or five years. So I've had a lot of look at those 50 to 60,000-year-old materials. Oh, yeah. Yeah, I'm working with Yonatan Saleh on stuff that's like between 160 and 200. Yeah, he's got a lot of cool sites down there in the south. I see him every time I'm at the museum. Yeah, he's pretty excited. He's also excited about getting out of Germany and South Africa. Well, has anybody applied to city and dating in Russia in the last decade? Does anybody know? Have they been in any preliminary studies or no one knows? Can you repeat your question? Well, I was asking, have there been any applications of city and hydration dating in Russia in the last decade? No. I've never heard of anything. No, not here. All right, maybe we'll try it. So our next presentation is going to be the Kozushima Obsidian Shuttle across the Pacific and the migration of Homo sapiens to the Japanese archipelago. This is by Nobuyuki Akeya. And we want to thank you for joining us here. I looked up the time in Japan, basically the middle of the night. So thank you for being here in person. Okay, can you hear me? Okay, let's get started. It's a midnight in Japan. I'm Nobuyuki Akeya, archaeologist of Meiji University. Meiji, my presentation title is the Kozushima Obsidian Shuttle across the Pacific and the migration of Homo sapiens to the Japanese archipelago. Before starting, let me explain the first photo. What you see here is the distant view of Kozushima Island from mainland Japan. The first island of the right is Kozushima Island, which is 55 kilometers away from the viewpoint. And the black current flows here, outline. My main purpose today is to present the archaeological evidence for the maritime transportation of obsidian during the area of a parliolic UP in Japan based on an approvinance analysis. Geological obsidian sources in central Japan. This figure shows the main obsidian sources in central Japan. Based on the X-ray flow sense, obsidian samples from central Japan were classified into eight geologic areas, including Kozushima. Kozushima is here. Kozushima, Amagi, Hakone, Tate Shina, Suwa, Wada, Wada, Takara-Hara-Yama. Obsidian Art Club on Kozushima Island. Kozushima is a small volcanic island located in the Pacific Ocean 50 kilometers away from the southern tip of Izupininshira. Izupininshira and Kozushima are separated by a channel with a maximum depth of more than 10,000, sorry, 1,000 meters. The channel has never been connected to the mainland by a land bridge. Even during the last Glacier Maximum. Kozushima Island offers five well-known obsidian outcrops, Ombasejima, Saunukazaki, Kannonura, Nagahama, and Sawajiri. This picture shows the obsidian lava flow of the Saunukazaki outcrop. Black-colored obsidian lava. Ombasejima is a volcanic reef located about five kilometers offshore of Kozushima mainland. We can see the obsidian lava in the crater, the sank under the sea. There was a vast amount of high-quality obsidian around the lava. Large-scale EDXRF provenance analysis. This analysis relied on the X-ray spectrometer shown here. The machine to the right has been running in my personal room. This is my personal room for more than 10 years. Now I am talking, sitting on this chair. I analyzed about 40,000 archaeological obsidian using these devices. Location of Mount Ashtaka. Mount Ashtaka is located between Mount Fuji and Suruga Bay. Volcanic ash and scoria from Mount Fuji were deposited on Mount Ashtaka and formed a gentle ridge on its southern side of this area. More than 19 Paleolithic settlements were formed on the ridge during the Upper Paleolithic. Discrimination Diagram. These scatter plots show the computational groups of geological obsidian in central Japan as determined by X-ray fluorescence. In 1994, Akihiko Mochizuki and I devised four indicators and discrimination diagram to identify obsidian. Based on these diagrams, geological obsidian from central Japan were classified into eight areas and 26 sub-areas. We started a large-scale non-destructive identification of obsidian artifact. Excavated from Mount Ashtaka and identified more than 30,000 archaeological obsidian. Amount of obsidian from different sources in the cultural layers. The photo on the left shows the stratigraphic sequence of upper members of Ashtaka-Rome. In which combination of black soil band and reddish scoria are repeated. This figure on the right shows the chronological phases of the Upper Paleolithic on Mount Ashtaka. Phase one. This figure on the right shows the chronological phases of the Upper Paleolithic on Mount Ashtaka. Phase one is characterized by trapezoid tools and H-ground earth trapezoid tools. The paragraph at the center shows the transition of obsidian sources by cultural layers. The blue color represents obsidian from Kozushima Island. Which numbered nearly 700 in phase one. Koz obsidian. Obsidian was the major risk flow material during the early EUP on Mount Ashtaka. This suggests the existence of water crafts and maritime transportation. Idemaruyama site. Idemaruyama site is one of the oldest EUP sites in Japan dating back to 38,000 years B.P. The upper photo is a panoramic view of the site located at the foot end of Mount Ashtaka. We can see the Suruga Bay. See Suruga Bay leading to the Pacific Ocean. The major, major risk tool of Idemaruyama site are trapezoid, trapezoid tools and point made from obsidian washale. Identification by EDXRF and verifying by NAA. The scatter plot on the left shows the identified obsidian sources from the Idemaruyama site. In addition, 26 obsidian artifacts were identified as being from Kozushima on Basijima. Considering the importance of these data, it is necessary to validate the result using another high-resolution method. The scatter plot on the right is a discrimination diagram for identifying obsidian excavated from Mount Ashtaka, obtained through neutron activation analysis NAA. A total of 15 artifacts from Ashtaka Phase 1, including Idemaruyama, were submitted for NAA. These sources were identified through X-ray fluorescence analysis. But before revealing the X-ray result, 10 obsidian samples from 10 different geologic sources, 100 samples in total, these are geologic obsidian, were submitted to Michael Glasscock at the Archaeometry Laboratory of the University, Missouri. Based on Dr. Glasscock's values, I created the scatter plot as the right side figure shows. The values of Ashtaka Phase 1 artifacts shown by red triangles overlap with the geologic obsidian, including the Kozushima source. The result of cross-checking on this scatter plot were identical to those obtained from X-ray fluorescence analysis, Kozushima obsidian shuttle. This picture shows the sea voyaging for the Kozushima obsidian outcrop by skin on frame kayak, skin kayak. Kozushima obsidian excavated in the Japanese mainland means that it would have made a return trip across the sea. According to the result of Provenance Analysis, obsidian transport continued for more than 4,000 years. For these reasons, I named it Kozushima obsidian shuttle. A local sea kayak guide, my friend Toshiyaki Shijima made successful voyages to Kozushima several times. This figure shows the GPS track data of his voyage in 1993. At that time, the branch of the colossal current, which usually flows from the west to the east, was very weak, but his kayak drifted to the west under the influence of a rapid tidal stream near Kozushima Island. Distribution of the Kozushima obsidian during the EUP. Kozushima obsidian has been identified not only at Mount Ashtaka, but also in the coastal area of central Japan. Please observe this map. The blue part of each circle graph shows the ratio and the quantity of Kozushima obsidian in relation to other sources. The sources. It is clear that the certain amount of Kozushima obsidian was transported to central Japan during the EUP. Migration of home surface to the Japanese archipelago. I will be talking about the holistic re-enactment project of voyages 30,000 years ago by Japanese National Scientific Museum and Dr. Yusuke Kaifu. Our dugout canoe left the Taiwanese east coast on July 7, 2019 here. After about 40-hour voyage, we arrived at Yonaguni-jima, the westernmost island in Japan. I designed this canoe and supported the crew from the escort ship. The expedition aim was to verify the means and routes of Homo sapiens migration to the Japanese archipelago at the beginning of the EUP migration route of Homo sapiens. This figure shows the routes and timing of migration to Japanese archipelago. The Tsushima route. This is widely accepted as the first migration route to Japan. And the Japanese upper-parioristic started 38,000 years BP. The EUP corresponded to marine isotope stage 3 and the sea level was about 80 meters lower than at the present, which is relatively high for the ice age. This is the EUP coastline. The Japanese archipelago is surrounded by the sea, Tsunaro Strait, Tsushima Strait, Uki Ocean. The Japanese, so the first migrant must have crossed the sea to reach Japan. The Idemariama site date back to 38,000 years ago, mostly overlapping with the beginning of the upper-parioristic Japan. Cosmo obsidian provides very important archaeological evidence for studying upper-parioristic. Furthermore, it collaborates the seafaring technology and ocean adaptation of Homo sapiens. This is Yonagunijima Island, Shiliat. Thank you for your attention. That's all. Well, thank you so much. That's fascinating. I'm going to have any questions. Dr. Ikeya, did you, so you made a replica boat from skin? No, I didn't. But my friend, Shiojima, made skin kayak. And yeah, I tried to ride and paddling it. Yeah. That would be a long, long trip on a replica boat. Interesting. So you're saying the 80 meter lower was 38,000 years ago. So it joined with one big island. Big island? You say Cosmo obsidian, Kozushima Island? No. If sea level was lower, 80 meters lower than present, Kozushima Island might link to neighbor island in this way. But as I mentioned, this channel was very deep and never connected even in the dwelling, even during the ice glacier. Was there continental glaciation then? Were there potentially ice, icebergs and things like that? Yes. I mean the LGM, Las Crecha Maxima. Challenging here. Are there questions? I have a question on your, on slide number 15 there. 15? You have this one? You've got a lot of sources from the island. And then there's a source pretty close by that has, there's hardly any obsidian from that source. Do you know why they would be going to the island instead of, you know, just back on land a little bit? You say why Parosik people cross the ocean, right? Right. Instead of going north to that. Yeah. Yeah. Yeah. Of course, in this region there is high quality obsidian around there. And many people ask me why tomorrow's habitat, they are to cross the ocean. It is very dangerous. But recently the Parosik people do the fishing. And I think yeah, some habitants live there. And sometimes maybe the major, yeah, they, they do fishing. So that they know the existence of cause of small obsidian. I think, I think so. Makes sense. Thank you. Maybe the other material isn't obsidian. Is it, is this comparison of different obsidians or is it all lithics? Another lithic material. Sorry. I was asking, I was asking if the pie chart is showing all lithics or just this type of obsidian versus other obsidian? Perhaps they're using local charts or something else. You asked a chart. I was just wondering if the pie chart, pie chart, doesn't have much obsidian. Is it because it's not, the other material isn't obsidian? Is this all obsidian on this graphic? Yeah. This pie chart, this pie chart shows obsidian ratio, not include other lithic material such as chart or sedimentary rocks. Only pie chart shows only obsidian. All right. Well, thank you. This a really great talk. Thank you. Thank you. So moving forward here, our next talk is entitled Archaeological Research of Krami Valley, Transportation of Obsidian in Neolithic Calculative Times, Preliminary Results of New Archaeological Research by Saba Joukati. So, just share my screen. Oh, you see the presentation. Yes. That's good. So let's get back to Calculus again. Today, I'm going to talk with my presentation about Calculus region, South Calculus region and its central part which played a great role in the development of late prehistoric populations and the archaeological cultures. Just to introduce, I'm going to show you most of you already know, but the significant obsidian sources in the South Calculus and that the very north of the many existing sources is located the Chikiani obsidian dome. The Chikiani obsidian is a very good quality, very black homogenous, but can be turned into reddish brown or even green variations. You see this picture. It's a modern quarry exploitation and huge obsidian blocks. So the obsidian appears at Chikiani as a segregation of in the one-metre-thick rhyolitic flows and the flows goes in the east. So these flows are dating around 2.8 to 2.9 millions ago by different analysis. You can see the obsidian on this spot, how it looks like and the landscape is quite changeable in Chikiani and in summer it looks like this and in winter it's snow cover more than five months. Just to show you how it goes the archaeological context in the South Calculus, we have a very dominant culture called Shomushula very culture. You can see the spread of the settlements. This culture was identified in sixties and then different international teams and archaeologists they worked on these sites. The sites are numerous, more than this, but I'm just showing you the ones which has been excavated and the publication. So these sites are spreading Azerbaijan in South Georgian and Armenian territories. Just to show you the quick examples, you can see the mud brick architecture, well-developed pottery technology, there's a numerous kinds of bones and stone tools artifacts, but also with high percentage of obsidian use and quality is really a big high level. To see in the smaller scale on the south Georgian territory, you see this green line indicates the south border of Georgia and you see the, if you see my cursor, so you see this Neolithic settlements, Shomushula very settlements and blue indicates the Calcolitic settlements. So these two chronological period doesn't have a much similarities with the archaeological material and so on, but the main thing which this population they have is the use of obsidian. Both of the cases we have a more than 95% deletic industry depend on obsidian and the use of Chikiani sources is important because the last three decades, the different scholars, they have made the sourcing of the artifacts from these settlements and its cluster together with Chikiani obsidian, which has that kind of fingerprint with the high barium and low chikunium. So it indicates that the area in the central part of Calcasus, this was published some 15 years ago by Professor Battalion. So these yellow lines indicates the area where mostly in late prehistoric populations they used mostly the Chikiani obsidian source. So we always were thinking about how was the possible exchange roles, exchange of these obsidian materials and for us it's really important in South Calcasus, the Khrami River, which is the biggest river in South Georgia and it looks like this picture and also the right tributary, Chokhiyani River, which connects exactly this link to the exactly Chikiani spot. The Chokhiyani River is looked like this and how important was this exchange process for that Neolithic population in the low length of Khrami River. So that's why we conducted archaeological surveys in the eastern vicinity of Chikiani source here in the circle and also an upper stream of Khrami River and this map indicates the digital elevation model as you see and just show you how is the difference between elevation. So the Neolithic settlements is located in the low lands and they have to travel to the highlands of Jabaheti Plateau. The difference is like maybe 2,000 meters or something like this and we wanted, so our goal was to find the Neolithic evidence in the Neolithic evidence in the highlands of Lesa Calcasus because we know that all the time they have traveled to this area but we don't have much evidence of it. So we wanted some kind of short-term occupation sites or workshop and so on. So the surveys were conducted in 2020 autumn. It was really hard with the pandemic but we managed to somehow work there and of course we have different periods, different chronological periods and the results but we're going to show you today some of the preliminary results of workshop, what we found and the rock shelter. Both are late prehistoric evidence of in the highlands of Calcasus, Lesa Calcasus of course. The workshop area is located in the northeast part from Chikiani source. It's around four kilometers and the position of the workshop is really important. We think that it faces to Chikiani source and also the Chikiani river valley which then connects to the main river. So it looks like this on the area. You can see on the distance the main Chikiani source and the left side the Chikiani river. As you see in the workshops but the hummus layer is kind of opened with a high percentage of rain in spring and so on and it is easily detected the different kind of chipping floors. You see this kind of ellipse form platform and there is a numerous significant number of obsidian debris, the flakes to broken tools and tools also. Of course there are also the materials from the Middle Paleolithic and Late Paleolithic but we wanted to find some exactly the small areas where it's a concentration of late prehistoric materials and we found this kind of conical ports and blades which is clear characteristic for the Neolithic culture. You see the Levo-Prussian technique of blades here. Also we found the stone chisel where you polished and worked. We think that it was used as an axe but then it was broken and then changed to the chisel. So because of these edges which go down and it's polished, very polished, maybe it doesn't seem like on the picture but anyway. So this is also the clear evidence of Neolithic in the area because these kind of tools they use only in Neolithic in the South Caucasus. We have also called the Levitic assemblage. All these materials are under processing and we are going to do some kind of traceology and it gives us more information of course. The scraper, the end scraper is really characteristic for a calculated period. Let's get back to the second rock shelter I'm going to share with you. The rock shelter location is in the upper stream of Khrami river as you see on the left bank of the river. The valley is on the highest point 18 meters and eastwards going down and we detected this rock shelter in the center of the valley. It's around 50 meters from the base of the river. Probably with the reason of falling boulders, the Azaltic huge blocks, the shape of the the shape you know it's quite a bit changed but in general picture you can see the spot and also the naked burnt layer we found there is around 40 centimeters. So this kind of short occupation sites doesn't have a huge, very thick cultural layer so this is really common for the area but so when we talked about the artifacts which was discovered in this layer it's quite tricky we can't say that it's exactly this period or something but we think that it's a late prehistory in general and in future maybe if we will work there because this sorry the site is under danger also because of the falling the boulders and also the villagers they always move on this layer so in quick future maybe we can excavate some part just to get the date or something you know so we want to do it quickly as if the pandemic will give us a chance of course. So the materials is really seems like misleading let's say because of blade broken blades you see here and also the one left blade so and also we found in the erosion soil two you see here two holes which was made bone and discovery of this hole also give us thoughts about the date because for example 500 meters from the from this rock shelter it was excavated in the 80s the Azzani shelter they had 20,000 of artifacts but none they have bone tools so maybe it gives us the information that it's a little bit later but it's really hard to say at this time so if we summarize all of these what we have here we have a Neolithic culture the Shumshula very very dominant strong culture and also we have these Calculated period sites and all these populations the contact permanent contact to the Javakheti Highlands and we don't have much as I told you much evidence of it but with these new results we think that it might be to cover all this all this gap so we will see we will continue to survey there maybe and to get more information with this point so you can we just thank for the funded organization and so on and the bibliography of used photos in the presentation thank you very much please give the questions if you have thank you so well one thing that that kind of jumps out to me is I wonder if you've considered using perhaps least cost analyses and overlapping where you find the sites and what you would expect through it to something like that yes we had we did it but the thing is that we don't had a good how can I say the so the the problem the problem was that we can't find the exactly how it looks like the the program needs more information you know to get you to get a good result let's say so first we we planned to make this survey and then we will do also this and then we combine everything and maybe so this is my phd some small part of my phd of course and maybe in future of course a good advice thank you no it's really interesting any other questions yeah I'm on all no I have one question yes it's only okay you should only one obsidian source in the wide range area you have just only one obsidian source in georgia yes in georgian territory only one we have yes you see here the the border but this is a huge dome let's say but we can say that this is one chikiani obsidian source it's a huge dome but we don't have a other spot let's say other here or something there no this is the chikiani which is very big but only one it's oba thank you good morning this is craig young um yes you see any secondary distribution off the highlands into them from the dome into the bigger drainage areas yes it's a good question because the khrami river which i which i showed you already has its secondary deposition maybe i want to quickly move on the yeah the because this chokhiani river as i told you takes the pebbles to the obsidian small pebbles to the uh to the stream and in maybe in paleolithic it was useful but for the neolithic people they don't care about these uh these pebbles because they needed good material right so for the neolithic and calcality people i don't think so it was useful the secondary source but uh in paleolithic of course yes maybe right thank you any other questions so i will stop sharing then well thank you salva very great so uh right now we're gonna take sorry break oh yes yeah i have a one question yes uh what about the hammer stones hammer stones yes we don't have any hammer stone yet from the workshop area because you know it's a numerous number of materials there and this material is quite mixed i mean because these obsidian obsidian and the territory around it was used from i don't know middle paleolithic to the iron age so it's mixed everything you need to really precise archaeological surveys to to find out the exactly spots you know it's really hard but with the point of of uh these hammers we don't have yet sorry it's it's really hard to distinguish also because uh you know you see this is it or not it's really hard the context is yeah i don't know and can can you distinguish between uh pressure and uh and uh percussion yes yes it can be done yes of course yeah and most of the yeah no yeah uh i want to say that the most of the neolithic neolithic i'm talking about the culture i showed you the show musul aleric culture most of the blades are made by by pressure technique for sure okay thank you and congratulations welcome yeah so we're gonna take a 20 minute coffee break here and move on then i will we'll talk about a timeline for the utilization of the carpathian obsidian sources uh by catlin t bureau so catlin went and take it away a timeline for the utilization of the carpathian obsidian sources most of the word obsidian sources are located along island arch or seaside opposed to this the obsidian sources in central europe are located literally in the middle of the continent looking at the palo geographical map of the region however we find that in the not too very distant geological past the upper myosin these conditions were met for the current area of the carpathian basin known as the paratetis ocean and later the panoneal lake the formation of the carpathian obsidians took place in those times dated between nine to 15 million years this continental position of the carpathian obsidian had significant consequences on the accessibility and archaeological distribution of these obsidians we must also note that the sources are relatively small and typically secondary the name itself is a bit misleading as the sources are not in the carpathians but in the carpathian basin more specifically in the tokai pressure mountains and they are also not related to the carpathian geological period which is lower myosin whereas the obsidian were formed in the upper myosin stage as it is well established name already we use the term as created by all the williams and colleagues carpathian obsidian are classified into three main categories the oldest highest quality and most popular being c1 or carpathian one on the territory of slavicia close to the hungarian border c2 hungarian obsidian is located at the southern and central parts of the tokai mountains in two regional centers around tochoa and mad at the vene these varieties can be distinguished fairly well on the basis of microscopic properties rarely we can come across obsidian of reddish-tint mahogany obsidian among the tochoa specimens this variety was also known and used though only in very small quantities the history of hungarian obsidian research dates back to the second half of the 19th century pioneering personality of hungarian archaeology florish roman introduced the subject into the focus of prehistoric archaeology presenting the first obsidian artifacts to the general public at the same time he promoted the collection of data by involving the general public landowners teachers even simple farmers he was implemented in bringing the eight word archaeological congress to budapest in 1876 where he presented among others the first archaeological obsidian distribution map for that time hungly as you cannot see the details very nicely I made a transcription of the sites that were marked for obsidian his efforts were followed by several researchers in the 20th century in the 1930s a wave of regional data collection took place in Poland by Kostrafski 1930 in Romania Martin Roszka in 1934 and in Slovakia Jan Szak 1935 the data collection was continued for several chronological periods in Hungary in the 1950s the introduction of archaeological analysis from the mid 60s onwards also resulted in further date points however nobody could aim at the regional integrity and coverage that was taken for certain within the Austro-Hungarian monarchy covering most of the archaeological distribution area of the Carpathian obsidians I had several rounds on collecting distribution data starting in 1981 and the last one completed in 2014 for the occasion of the 2019 obsidian conference in Szároszpátok I have collected all chronologically meaningful information within Hungary starting with the paleolithic distribution data emphasizing the most important utilization period the Neolithic and also presenting some of the late uses of Carpathian obsidian by the half of C14 based graph constructed for the volume Hungarian archaeology at the turn of the millennium I would like to show a timeline for obsidian use regarding the Carpathian obsidians in the lower paleolithic period there is only one site of authentic information known that is Véter Szöldős it is in Transnanubia meaning to the west of the venue very far from the obsidian sources and uses basically local pebble materials no obsidian whatsoever we have no evidence that the source area of the Carpathian obsidians was inhabited in the lower paleolithic the few hints of maybe lower paleolithic obsidians cannot be accepted in the yes just to mention that Véter Szöldős is a really very complex and beautiful site with remains of prehistoric mantle in the middle paleolithic period again we have no information on people living at the source area for C1 or C2 obsidians the relevant cultures are always shown within the chronological scene obsidian however is known to occur at several sites quite far about 100 kilometers from the sources in the framework of well-dated mustinian sites that is Szubaljuk, Kecskézgwaja and some other minor localities in the region of the C3 sources we have evidence of local workshop activity on the source at the same time we have no information on the distribution of C3 obsidian on archaeological sites at the chronological boundary of the middle paleolithic and the upper paleolithic we have a few very important sites with obsidian we go back to the former site to see Pilisanto II Pilisanto II is a cave in Transnagubria with one maximum fauna and indifferent archaeological material the single obsidian flake however is the oldest known obsidian item to the west of the Danube about 300 kilometers from the source the early upper paleolithic industries Salatian and Orinacian both new and utilized capetian obsidian also in Slovakia even more when the source region was already inhabited in the Orinacian times these are the relevant cultures and let us mention one small site legend Kanditanya which can be dated to early upper paleolithic probably Orinacian this is the first site when all the four types of obsidian capetian one two varieties and even Mahogany obsidian occur together yes it was also used in some very interesting leaf point industries of early upper paleolithic character in the environs of Mishkoltz and cave sites around the city yes the late middle paleolithic gravity and culture utilized C1 and C2 sources as well the highest ratio for obsidian was observed in Hungary at Modrokka Rasturhenye about 20% of the lithic industry abounding in C2 local types for the contemporary large workshop site Orca obsidian was a non-local material it is interesting to note that the first Mahogany obsidian was spotted here most of the red obsidians were observed at the gravity and sites as most obsidian varieties C1 and C2 occur together with post volcanic limbic silicates we can often observe that the two kinds of raw material travel together limbic silicates in Hungarian geological literature limbic core sites are excellent raw materials for use but rather difficult to work on due to many plant for sales found in them the highlight of obsidian news for the Carpathian obsidian was obviously in the Neolithic period early Neolithic cultures it is interesting to note that the early Neolithic communities especially the Kurdish culture and the earliest phase of the linear pottery culture most new and utilized obsidian and in relatively large quantities in this time the source region was not populated so the access to the sources was probably realized by seasonal expeditions opposed to this in the middle Neolithic the LBC descendants Tissadeau and Buccaultures used to live very close to the obsidian sources most probably new and excellent obsidian sources were found and exploited and probably depleted as well this is indicated by the appearance of large blade cores yes that took part in long-distance trade as well also in the region local workshop sites at the cash of Trepegove and the former was the well-known obsidian depot found close to the Romanian border near Lugos after the Buccaultures the immediate neighborhood of the obsidian sources was that populated and minor inferior sources were in use probably by expeditions and raids yes this is still belonging to the middle Neolithic horizon workshop sites around the obsidian source areas here we are at the late Neolithic and most of the activity is related to Lenya culture which is very well known for its industrial efforts to mine and to work from it yes on the territory of Lenya culture of Trans-Anubian origin a clear advance to the obsidian sources can be observed whereas at the Trans-Anubian homeland certain sites seem to operate as local distribution centers one of them is the site Saka in Sehei the most typical element in the Turkic world micro blade cores and micro blades we can observe the longest distances of archaeological obsidian distribution in this period in the Copper Age projectile points appear they are typically found in funeral context they are also characteristic of the first part of the Bronze Age again a special funeral site of the pin grave culture the importance of the stone is slowly decreasing but they are definitely present on their own right till the late Bronze Age this example that I show here is a notable exception from Chograd Kettőshalom for the stone artifacts of the Iron Age we have to consider secondary use as say fire flint yes the pieces come again from graves of the Skitian Late Iron Age context in modern times there is a growing interest in obsidian by collectors tourists and jewelry producers and buyers that make sourcing a challenge especially if experimental napping is involved it is not the subject of the current talk but we have a growing set of analytical information on Carpathian and other European archaeological obsidians whereas C1 and C2 obsidians often occur together we have certain instances along the borderline of the Carpathian obsidians where the site was supplied from several sources typically C1 with Lipari obsidian on the southwest and C1 interacting with Malian obsidian on the southeast hopefully in the near future a lot of new information will be available by the publication of the last obsidian conference in 2019 and there this paper is very much a preliminary report on the analysis pxrf analysis of obsidian from a late Neolithic site in southeastern Hungary and I will attempt to pronounce the name of the site but county bureau is probably going to throw a fit if I get this very wrong but I think it's pronounced but we're going to say we're going to call it gorgeous for short so this obsidian provenancing study is really just a small part of an ongoing multidisciplinary research project into the provenance and the typo chronology and technology and functional analysis of lithic raw materials that were used at this particular site and if I go back one slide seem to have advanced right the people involved here friends who are that of course is the excavator of the site from which the artifacts came the all the hard work on the the lithic artifacts has been done by Elizabeth Estanini and Barbara Wojtek and my role has been a very small one just analyzing with pxrf the obsidian artifacts now the site itself if you can see my cursor here is represented by this red dot here it's at the confluence between the river Tisa and its eastern tributary the Marosh which starts off in the Carpathians in Romania and so the site is within the great Hungarian plain now perhaps just to say the great Hungarian plain lies within the Carpathian Basin and if you look at the lower map in this slide this is the Carpathian Basin a lowland area mainly surrounded by mountains by the Carpathians in the the northeast and southeast the Alps over here in the west and down in the southwest you've got the beginning of the Danaric Alps which continue all the way down the western Balkans now today this is very rich agricultural land back in the Neolithic it was a very different environment in its natural state it was a flat lowland area or primarily steppe or prairie grasslands broken by swamps shallow lakes and patches of woodland along river flood plains so it's a very different environment before regularization of the river's drainage of the area there's a lot of water around in the the great Hungarian plain now during the late Neolithic the the Carpathian Basin um was populated by a mosaic of archaeological cultures and over here in the in southeastern Hungary and the great Hungarian plain you have the Tisa culture which is the the characteristic late Neolithic culture and in the latest stages of the Tisa culture there evolved a series of regional groups so that you have the classic Tisa culture from just the west of the river Tisa extending eastwards into the the lower parts of the drainages of the Kurdish river system and the Marosh system and to the east of that in the upper Kurdish you have the Habai culture and then to the north in the head waters of the river Tisa you have the Shysalom culture now you can imagine that in this region which was where there was a which was a lowland area covered by fairly recent sediments of holocene implies to see age in the main obtaining good quality lithic raw materials for tool manufacture was a problem for any Neolithic communities and just for the benefit of those not familiar with European prehistory when i'm talking about the Neolithic i'm talking about the period of the first farmers and the the Tisa culture people were not the very first farmers on the Hungarian plain they were fairly well established farmers about a thousand twelve hundred years after farming first reached the Hungarian plain but securing supplies of lithic raw materials was a bit of a problem the nearest sources primary sources are at least 60 kilometers away from the Georgia site and so the raw material had to be brought into the site either in the form of unworked stones or as ready-made artifacts so I meant to advance the slide but the point about this slide this is by Elizabeth Estanini is that okay we have some obsidian in this slide but there's a lot of other raw materials that were being other lithic raw materials that have been used here there was various kinds of flint and it was coming from considerable distances away and the obsidian is really only a small proportion of the overall lithic assemblage and that's a very important thing to grasp about the Neolithic around these in this particular area of Europe so this is Georgia two views from google earth and the three-dimensional view over here or the street view where you can see that it's actually a settlement mound it's a tell rising about four or five meters above the the floodplain of the Tisa and it's quite a big site it covers approximately five hectares and for Ensorbat's excavations which I think to place between 1978 and 1996 examined about a thousand square meters of the site and obviously about several meters depth of sediment so it's a major project and his excavations reveal the late Neolithic sequence spanning the middle and late stages of the Tisa culture and radiocarbon dates span from about 4850 to 4500 BC Cal BC so you basically had a settled community here that remained there and developed there continuously over a period of several centuries and so just as you have in the Near East and Southeast in Europe you get tell settlements developing now the other thing to note about this slide if you can follow the cursor just in the lower part of the slide you know goo the google earth photo picks up these old river features these old river channels so you can see that the river channels are shifting about through time you get seasonal flooding or you had seasonal flooding in this area and they were constantly shifting bodies of water now because raw materials were having to be acquired from some distance away and it's likely that they were being acquired through participation in exchange networks through cultural connections it follows that scientific analyses to identify the raw material sources are a very important step first step in reconstructing any exchange systems that operated on the great Hungarian plane during the late Neolithic now we've had Katy talking about this but the nearest obsidian sources to Georgia are on the edge of the Carpathian the western Carpathians to the north and we essentially there have three source groups we have the Carpathian three source group in the Ukraine the Carpathian one source group in eastern Slovakia and the Carpathian two source group in northeastern Hungary and as Katy mentioned you can distinguish visually in many cases the obsidian that comes from these different sources so typically the Carpathian one obsidian which is very high quality stuff in the main it's typically highly translucent has a very glossy appearance and in some samples it's often speckled with black or dark grey speckles or even stripes and bands so it's very distinctive type of obsidian Carpathian two generally speaking is visually rather different it's less translucent characteristicly black or dark gray in color and usually only slightly translucent to the edges in when you see thin pieces and there are two subtypes but I would hesitate to Carpathian 2e and Carpathian 2t which we can distinguish by XRF but I would hesitate to try and distinguish them visually and then in the Ukraine we have Carpathian three and Carpathian three it's not one of my pictures down in the lower left of this slide it's black it's relatively glossy and you get these rare white inclusions in it normally plagioclase and as you can see from that example has a slightly hackly fracture so it probably wasn't such good material for napping I've never tried it but I assume it wasn't and it wasn't as popular in prehistory in the prehistory of central and southeastern Europe as the Carpathian one and Carpathian two and I'm just going to cut off my um if you'll bear with me my WhatsApp now oh before I say that the the straight line distance between Gaugia and the Carpathian sources you can judge by the scale is on average I suppose between about 220 and 260 kilometers now of course the work at Gaugia the obsidian characterisation work is by no means the first characterisation work that's been done on sites on the Hungarian plain and even in the late Neolithic sites there's been a lot of previous research some of its involved lab based techniques like neutron activation and prompt gamma activation analysis um which you lose usually looked at small samples of material from these sites and then later on there've been some pxrf studies and we can discuss of course uh ad infinitum the pros and cons of lab based versus pxrf but the great value of pxrf of course is it's extremely flexible low cost uh you can take the instrument uh to the actual sample and more important than that because it's quick because it's cheap you can analyse more and non-destructive you can analyse every piece that you have from a site and that's perfectly possible because in sites this far away from the the sources the obsidian rarely forms more than a small proportion of the assemblages and it's perfectly possible to analyze every piece on the site now there was a very recently a study by daniella riba and i don't know if that's how she pronounces her name i think she's based in the state uh although she has a german sounding name and she published an article recently um in which she examined and she analysed by pxrf um samples from the hapai culture and the tisa culture and she analysed in total 203 artifacts from seven sites uh nearly all the materials she looked at came from the carpathian one source there's nice translucent banded uh stuff which is very good for napping uh but she did find some carpathian two material not very much she found carpathian two she found one piece at each of four sites um what was interesting is that two sites that belong to the tisa culture down here uh the carpathian two material came from the twoe source in hungary and at two sites that belong to the hapai culture they came from the carpathian two t source so um not unreasonably she assumed that this was culturally determined and that the hapai culture communities belong to um one exchange network and the tisa culture people belong to another or to a set of networks now we'll come back to that in a little while now this is not daniella reber i was just making the point that daniella reber used a brucatracer 3 to do her pxrf analysis uh i was at that time using a night on xl3t instrument uh has similar capabilities to the brucatracer it's not the latest generation of analyzer and neither of them are but it does the same job effectively uh you it's less flexible than the brucatracer in some ways but it's much more user friendly it's a great instrument for actually training students not to use because it's so easy to use but now coming to the uh engorger assembly itself and the pxr analysis that we performed there um we only began this project we only were able to make one visit to the uh to the museum in segad over two days and that was in august 2019 when i was able to measure 131 out of i think a total of 386 pieces of obsidian so roughly a third of the obsidian assemblage from the site and on that occasion just as a first step i prioritized the larger thicker pieces to avoid getting problems of thin samples or samples that didn't cover the measurement window which can lead to distortions in the results and i was to do a follow-up visit at the end of 2019 but we all know that the covid pandemic started around that time and that put pay to any further visits so i hope to get back there this year but um we'll see what happens now the results well very very briefly it's very obvious this slide uh of the 131 pieces that i analyzed uh 126 pieces uh could be matched very clearly to carpathian one and the the ellipses here represent the range of variation in geological samples that were previously measured with the same instrument and so you've got 126 dots within that carpathian one ellipse there so very clearly the vast majority of the stuff at gorge is coming from the carpathian one source now in the one third of the assemblage that i analyzed there was no carpathian two or carpathian three but we don't expect carpathian three and down here anyway but what i did find was several pieces that i couldn't match with any known european or anatolian obsidian source now that's not a problem uh or what was the problem was to save time i didn't measure the light elements on the instrument because that would have added two minutes measurement time for each sample so i left out the light elements which meant that i didn't measure silicon but i didn't manage to measure one piece right at the end of my visit um that one piece um and this is just a rough estimate it gave um an sio2 content of over 80 percent which would be unusually high for obsidian and certainly for carpathian obsidian so i'm guessing that the four or five pieces that i couldn't i couldn't provenance are probably not obsidian at all but some other salacious material probably a high quality form of chert or flint that just happens to be black and shiny but we'll determine that uh when we go back so to put all this into some kind of broader perspective i think we have to consider the wider regional and chronological context here now the dotted line here that i've drawn on this map here represents the eastern and southern boundary of the great hungarian plane so you can see that the great hungarian plane is not confined to hungary it extends into the neighboring countries it extends into the western part of romanium all the way down here and quite a bit of it is in northern syria the boybord in the region and on some definitions the great hungarian plane it extends into curation and there's been a lot of obsidian provenancing research done recently in romania by not just by me and my Romanian co-worker at Enambaranance but by other people um um Florin Drashevan working with Michael Glasgow um yeah Paolo Biasci has done work down in this region here and the area known as the Banat which is the biggest part of the hungarian plane that falls inside romania but i just want to focus on two areas where i've analyzed obsidian from maybe about roughly 20 30 sites in these areas ranging in age from early upper panolithic through to bronze age so up here in the northwest which is relatively close to the obsidian sources this is the wash region of romania and down here within the Banat uh i'm calling it the Arad plain the city of Arad is somewhere here and we're on the Mooresh plain the the Mooresh or Marosh river valley not very far from Gorgia now what i found in measuring obsidian there was that you know in this wash area up here um the back in the upper panolithic you find the use of all three obsidian source groups even carpathian three not much of it it's only there in the orignacean probably around 40 000 years ago everything else is carpathian one and two and then after that you don't see carpathian three in this region it's fairly localized its distribution is fairly localized around the source um but generally speaking between these two regions there is a marked change that takes place after the early after the first few centuries of the early Neolithic around 5800 bc before 5800 bc uh you find carpathian one and carpathian two being used so the very early Neolithic people here are using carpathian one and carpathian two then in the later part of the early Neolithic and then onwards into the late Neolithic from the Copper Age and then the Bronze Age you don't see carpathian two anymore you only see carpathian one um if you go further east in Romania you get the same pattern down into Bulgaria you get the same pattern down into Serbia you get the same pattern you get this change after about 5800 now interesting question why is there carpathian two obsidian in the late and Neolithic of southeastern Hungary but not in western Romania and I think the end I don't know the answer to the question to my own question but I think our colleague Georgia actually touched on a possible answer to the question uh in that in obsidian can be transported not just by people it can also be transported by natural processes and one obvious natural process of rivers and that we've got to remember that these carpathian obsidias he's not lava flows it's his pyroclastic material it's little volcanic bonds and small pieces that stuff can easily get transported down rivers and end up in river gravels so it's not beyond the bounds of possibility I'm not saying this is what happened that uh some of this this little bit of carpathian two obsidian we find down in the southeastern part of the Hungarian plain might be being picked out of river gravels um you know so that's something for the analysts to bear in mind when they're examining the size of the pieces when they're examining the cortex but there are alternatives now how does that explain carpathian two here and not here well simply because of the way the rivers flow the rivers come down here there's no way you're going to get obsidian a river transporting obsidian from here are you into the wash or into uh or up the Moresh valley just not possible so if you're going to have river transport of obsidian yes you'll get it down here which you won't have carpathian two obsidian you'll get it down here but you'll never get it over here so final remarks on this um as I've said uh our provenance studies of the Gaussier obsidian are very much a work in progress and uh the other thing of course we discover is that not everything that's black and shiny uh is obsidian uh so you can get other materials that look a little bit like obsidian and small pieces and turn out not to be when you do the geochemical analysis um the other very important point I think is when you're getting you know 50 or 100 or kilometers or more away from these sources obsidian is only one of a vast range of salacious rocks that were being used by Neolithic people and it's you know and they were not uh to them it probably wasn't a particularly special material it was just another salacious rock and it's not until really we get into the Calcolithic period in this region that we we start to see obsidian being used uh as a special material it starts to crop up in Calcolithic burials rich Calcolithic burials like the Varna burials with all the goldwork when you get something similar in northwest Romania in the Calcolithic with goldwork copper and nice arrowheads made out of obsidian so we always have to bear that in mind and I think the other important thing about this is what pxrf has allowed us to do is to analyze every single piece and I think that's extremely important when you're studying these sites so I'll end there and thank you all for listening I particularly like to thank my co-authors who did all the heavy lifting on the lithic analysis here so thank you very much well thank you Clive any questions for Clive I want to ask you once a quick question thank you for interesting do you hear me yep thank you for the really interesting presentation I just want to ask you about the management of producing tools because do we have the sign that the materials were produced on the settlement or it was totally produced in on the on the spot of the source and then it was brought the materials like like a half ready or something okay well I think that's a question from my co-author Elisabeth or if she's with us or maybe okay hello yes I think that the material was chipped on the side since we have several indicators of this like nodules, corticated nodules and pieces with cortex and debitage wastes so it was processed at Gorge okay okay do they had a specific place where they did it or the the materials is around I mean the the rest of the of the chipping process is around or do they had a special place for it no it is at the almost in every type of units and features like bits houses refuse bits so it's everywhere okay thank you very much together with other raw material so we didn't find a segregation area with the only obsidian chipping activity thanks thank you very much all right well thank you and of course Katzlin do you want to go real quick just a quick comment to Clive I have also found one piece of obsidian from the site had by which doesn't fit into anything it is obsidian it is not carpathian one two three four whatever I don't know the obsidian okay has it been analyzed by yes it was analyzed by energy dispersive spectroscopy in scanning electron microscope okay it should be equivalent to something like x-ray well it's possible that some of the pieces that I couldn't source to the carpathians are indeed obsidian but you know not having done the light elements on most of them that wasn't different because we also had some silica strokes that were mistook for obsidian but that's not the case here so just add to your list and this enigmatic source has one more from her by okay good it can be some further work to compare and try to to see if they are matching I mean and look for a missing source all right everyone welcome back we are moving out of sort of the regional study of things to more sort of theme-based talks the first is analytical methods in the afternoon that'll bring us through the rest of the day and then we'll start a new or another theme tomorrow morning the first paper is a pre-recorded one titled chlorine sodium dioxide as a geochemical index of obsidian aging by franco forest first martin and enrico massaro chlorine to sodium oxide as a geochemical index of obsidian aging a preliminary report by franco forest martin and enrico massara in a previous paper forced a martin rotolo nazare and carapazza 2020 proposed a new geochemical method to determine a provenance of the four italian obsidian sources widely exploited during prehistory matty arsi palmarola leapari and puntale area the method is based on the amount of chlorine a minor element in igneous rocks with average concentrations of a few hundredths of weight percent chlorine has an enrichment of an order of magnitude in rhyolites volcanics with a high silica content is detectable with the most common and low-cost analytical method for example semidepma table one and exhibits a good quantitative differentiation between the italian obsidian sources the solubility of chlorine and magmas increases with the content of network modifying cations lowlands turn 1994 especially sodium an element that also exhibits a good quantitative differentiation between the italian obsidian sources labordine cattalia 2006 and 2010 tycott 2002 forced a martin italiai 2020 verified the efficiency of the chlorine versus sodium oxide scatter plot to determine the provenances of obsidians by testing an adequate number of geological and archaeological samples clusters in the scatter plot of figure three discriminate unambiguously the four italian obsidian sources and also some subsources this study we are investigating the geochemical relations between chlorine and sodium in the obsidians of the italian outcrops taking into consideration the various subsources of the volcanic systems that have generated obsidians table three reports chlorine and sodium oxide concentrations and weight percent and the relative proportions of both elements for each source the last column gives the ratios of chlorine to sodium oxide the ratios of chlorine to sodium oxide decreases according to the sequence lepari ponte laria palmarola and montiarcy this sequence follows the growing formation age of the obsidian outcrops one can ask to what extent is the chlorine to sodium oxide ratio correlated with the formation age of the considered obsidian outcrops due to difficulties in performing radiometric dating on volcanic glasses the literature sometimes reports approximate and contradictory data in table four we included those obsidian outcrops for which the literature offers reliable radiometric dating the mean values of chlorine to sodium oxide ratio relating to the five selected obsidian outcrops appear well correlated with the radiometric ages a decreasing trend is evident indicating a higher depletion rate of cl data indicate that the chlorine to sodium oxide ratio r is a slowly decreasing function of the time we suggest using it as an age indicator of obsidian sample formation let's consider the timescale tau of this decrease that is the time in which a significant change typically by a factor of about two of r is produced this time can be defined by equation one it can be either constant or depending on the value of r the case of a constant tau corresponds to an exponential decrease as shown by equation two where the value of the chlorine to sodium oxide ratio at the epoch of the formation of the obsidian sample is indicated by r zero another interesting case corresponds to a time scale that increases for a decreasing value of r which implies that the decay of r is fast for obsidians with a high content of chlorine and becomes slower and slower when r reaches low values the simplest cases to assume that the inverse of tau is proportional to r as given in equation three the resulting function r t is the hyperbolic law in equation four where the constant a is the inverse of the initial value of r we computed the best fits of these two laws to the data by the minimum least squares method and estimated the parameters values the same formula is used in geophysics for describing the decay of aftershock frequency after an earthquake and is known as the amore law best fit values are written in the formula of figure five where the dashed green line is the exponential law the red solid line is the hyperbolic law the two laws describe well the aging of the obsidian samples but the hyperbolic one is slightly closer to the data points than exponential with a reference to r we have estimated an error of about five percent for its higher values and ten percent for the lower ones we applied our age estimator based on r for evaluating the ages of some Italian obsidian subsources not yet radiometrically dated we know from the literature that le par canado dentre subsource was formed in an eruptive epic proceeding that of gabalato and estimated between 20 and 9 000 years before the present for an yet ally i 2013 our results for le par canado dentre indicates about 21 000 years figure six the monday rc subsources named sa sp1 sp2 and sc according to literature have ages between 3.6 and 3.2 million years before the present mattanini and villa 1993 big i see at ally i 2005 according to our estimates the sa subsource could be younger about two million years while the ages of sp1 sp2 and sc are compatible with those proposed by mattanini and villa 1993 fig dot six we applied our curves also to estimate the age of obsidians from the greek island of mellos using the chemical data of case tovsky at ally i 2019 relative to a group of eight obsidian samples mean value of r equals 0.016 age estimates reported by literature for mellos obsidians are quite different dorani at ally i 1971 between 8.95 and 8.35 million years aria's at ally i 2006 about 1.6 million years our exponential law figure seven green curve gives an age not consistent with both the two estimates of literature while the hyperbolic law red results in a very good agreement with the age values of dorani at ally i 1971 yeah concluding remarks we put forward the hypothesis that the r decreasing as a function of the obsidian age depends on a number of alteration processes that the volcanic glasses do it seems that these chemical and physical modifications act with comparable times of occurrence for the different obsidian outcrops despite the different initial conditions of the progenitor magnus to test if the decay of r through geological times could be used as a geochemical proxy of radiometric dating we intend extending the curve calibrators carrying out on each sample both are in radiometric measurements testing of the law validity for samples out the Mediterranean area evaluating the decay law and its dependence on initial values of r particularly for rather young outcrops any contribution to add new analytical data relating to obsidian samples of known age will be greatly appreciated thanks for your attention the authors will be happy to receive your comments and to answer your questions email sedarius at rocketmail.com okay so i encourage you to um to reach out to the authors if you have comments or questions about their work the next paper is going to begin a little early which is my responsibility so i'll go i'm ready to go now and uh so i'm how's that can everybody see that okay yes obsidian objects supplying the malinubis distance statistic internary diagrams with r it's myself lucas johnson with kyle brind kathy davis and darin duke historically xrf analysis that thicker objects maximize fluorescent sufficiency for mid-z elements such as strontium through zirconium the thickness threshold of four millimeters is often cited as it is the point at which incoming x-rays all incoming x-rays are absorbed by the sample for obsidian zirconium in obsidian matrix however in practice 2.5 millimeter thick samples can use similar part per million values as kathy davis and others have shown parts per million become skewed when objects are at at or thinner than 1.5 millimeters in thickness the skewed and overestimated parts per million are typically caused by normalizing to the cotton scatter peak or some other physics-based parameter of the calibration for example jeffrey fergerson in 2012 illustrates this issue by overlapping samples of very thickness here we show glass mountain samples with very thickness the the thicker sample shows higher elemental peaks and confidence scatter as well as higher count rates per second or what we turn what we use is valid count as it is known in the brooker software these samples and all the samples in the presentation are analyzed at 40 kb for 240 seconds using a tracer 3 sd here is an example of glass mountain obsidian with very thickness for illustration over here to the right there's three three different spectra following factory recommended calibration procedures comparisons of ppm for these three samples from the same rock show significant differences in at least three elements this elemental difference is caused by the competent peak normalization and results in certain elements being either under predicted or over predicted as sample thickness decreases these issues can affect a confidence sourcing of shatter fragments and most pressure flakes unless data are transformed to ratios or relative peak percentages the exclusion of these data in sourcing project can significantly affect archaeological interpretations of mobility and tool manufacture or retouch taking place at great distances from obsidian sources the reason for this lack of confidence or the the use of ratio data is that source library samples consist of larger thicker thicker samples termed infinitely thick for thick source library samples valid count is highly clustered in distribution in this clay in this case those samples with thicknesses at four millimeters or greater return valid accounts between 3,000 and 5100 counts per second again here valid count is a proxy for sample thickness a recent project in the western great basin some 800 artifacts were analyzed their count rates per second range from 218 to 3550 this distribution this distribution alone suggests ppm will be highly skewed when compared to a thick source library the use of ratio or other transformation is the default sourcing method this slide shows the range in valid count for 400 vody hill specimen part of a simulated bipolar reduction exercise that produced more than 12,000 flakes all of which were measured for length width and thickness and analyzed with edx ref experimental study reduced cobbles to produce a variety of core reduction and retouch flakes as well as more small shatter for the body hill obsidian show a noticeable skew in most parts per million compared to the results of larger samples so a review of the literature on sourcing small artifacts archaeological flakes finds the use of both calibrated part per million and raw element peak counts the latter of rich are often term semi quantitative these two sets of data are commonly explored with any x or f unit and can reinforce one another in sourcing studies so two major study two major studies demonstrate the usefulness of data transformation the first huge 2007 describes the use of peak intensity ratios to source small artifacts his ratios are listed here in his publication Ellery Fram also outlines the use of part per million ratios and are provided of strengths in his article a couple of important distinctions appear in these approaches first some North American sources do not contain strontium or low or low amounts of it so in order to use it as a ratio values must be relatively high above the detection limit in other words part per million must be higher than one to five part per million fram on the other hand working in the caucus region shows that strontium values for regional sources contain about 20 ppm or higher in either case the methods and ratios put forth do help to resolve sourcing small flakes or microdebit charge it is important to note that peak element peak intensities and by extension peak ratios are instrument driven and therefore specific to the instrument and likely not comparable between labs ppm ratios on the other hand can be reproduced and use with those other statistical methods as fram demonstrates with discriminant function analysis deep or dfa this hyperspace method is a robust multivariate analysis that effectively enhances what can be seen in bivariate or matrix plots using two or more ratios to further source microdebitage robert jack jack and hyzer 1968 was was the first to display relative peak percentages using ternary diagrams with rough outlines of various sources used 2010 much much later offers a more refined use of relative elemental percentages displayed in ternary diagrams and he targets these three elements because they are adjacent to one another on the periodic table and respond similarly to variations in specimen thickness here you can see strontium zirconium and rubidium and these plots smaller effects are plotted with dash ovals representing the range of variation of a given source these ovals are inferred from extensive knowledge of how particular sources vary based on thickness and diameter of note these three elements are normally used to discriminate many california sources thus they may not be effective elsewhere but it is easy enough to substitute other mid z elements many statistical programs offer the ability to plot relative percentages using ternary diagrams as here as you can see here as well as we've seen a few other presentations but more robust statistical methods that require range of variation require additional steps a ternary diagram often referred to as the simplex in geological sciences is a useful way to visualize relative contributions of three variables when the three are summed to 100 in geology it is often used to determine the relative percentages of felspar and rock compared to other mineralogical components in geochemical studies ternary diagrams take the sum of three elements and determine the percentage of each relative to the sum and plot each one on the side of the triangle axis in order to apply confidence statistics in a ternary diagram that reflects smaller artifact sizes the problem becomes how to replicate this variation in a source library in other words what is the base data required to capture small flake variation that can also be reproduced and applied elsewhere if necessary here's an example of 24 sources plotted on a ternary diagram with only large source samples you can see the relatively small confidence regions depicted for each source with the introduction of small pressure flakes or shatter taken from each larger sample the region the confidence region of variation increases significantly the introduction of small flake samples shows a bimodal distribution by valid count this is a distribution some sources here you can see the distribution but there's not an equal complement of small and large in this distribution including relative peak percentages for small samples calculates a larger range of variation that's represented by the 95% confidence region confidence regions in ternary diagrams also have a history in the geological sciences and various methods for their calculations have been described here you can see a list of those references most of the past and recent literature uses a modified malonovus distance statistic or our d statistic and may require extensive calculations to compute depending on how many variables are included the use of the d statistic is commonly used in geological studies to determine outliers and multivariate analysis from a calculated centroid uh citing glasscock 1998 and 2019 the resulting computations show a modified z-score that is not related to the the Euclidean unit of measure in other words a concentration of zero to n a z-score estimates how many standard deviations a sample is from the mean for a given element but the d statistic provides a non-euclidean variable value expressing a distance for central tendency programs such as jump by sass provide an upward control limit as you can see here in the bottom of the slide to determine the limit 95 variation of it and any samples above this line are considered outliers the issue with the type of d statistic is that it limit the limits move with additional with addition and traction and samples from a source population the d statistic in obsidian sourcing is often coupled with discriminant function analysis to ensure that there is no overlap between sources in a single geographical region once the d statistic is minimal or the outliers have been removed the discriminant function analysis is around to show no overlap between source groups all of this is done prior to sourcing artifacts in this example the queen obsidian source in western nebada three of them are confused with the mount hick source using rubidium through niobium so we would want to go back and inspect these using recent case study confidence regions were calculated in r with the gg term package and other dependent applications such as gg plot 2 the code is straightforward and can be expanded for visualization or addition or additional groupings the basic data include various source groups with relative elemental percentages already calculated different from geological studies of relative mineral concentration its application to obsidian sourcing requires thorough evaluation of geocomical groups and a complementary distribution of small and large samples it is important to note that this method is not intended to substitute for the use of ratios described earlier but the supplement if necessary and to expand on earlier work of jack and hugh's enabling a reproducible procedure for calculating statistical confidence regions using three variables displayed in two-dimensional space in some instances dozens of sources can be plotted to reflect geochemical resource map although sources with similar geochemistry will overlap many sources from northwestern nebada situate in the lower left corner and must be replotted using element ratios to separate them further work in north central california for example examine hundreds of artifacts as part of a larger water management project area is at the confluence of multiple indigenous groups that resided in the central valley as well as in the uplands of northwestern nebada multiple residences were investigated where tool retouch occurred thus most samples were of small pressure flakes and other retouch flakes here's the figure showing that many of the small specimens emphasize california obsidian sources as well as those near the modern they they states of nebada in california here you can see the need to use other ratios to separate buffalo hills boaty hills and south waters more confidently another substantial sourcing project occurred within the area of northwestern nebada during investigations to restore freshwater springs during the analysis of the x or f results multiple obsidian source regions were identified in order to illustrate their regional diversity attorney plot was used here two major clusters of artifacts are shown in the center is a group showing sources local to archaeological sites as well as sources imported from northeastern california the other group in the lower left hand corner plot includes the grouping of three major north western nebada sources while overlap is present in the ternary plot and the three north western nebada sources appear as one group the addition of bivariate plots show a clear separation in this example the use of peak intensity ratio is necessary for the separation in these peak intensity by bivariate plots 95 confidence ellipses are shown for both large samples and those with large and small samples obtained you can see that how the ellipses changes large and small samples are represented by the dash line in summary sourcing small artifacts requires the transformation of ppm or peak intensity data these transformations minimize the difference in element concentrations between those and fifth in thickness and those with suboptimal sides important transformations include geographical specific ratios and may include the need to determine the relative percentage of at least three elements but even with these transformations some very small thin specimens may exhibit more variation than equally transformed thick reference materials here we have shown a procedure to introduce an equal complement of small and thin source reference material and using these additional samples in combination with a function within our package of gg tern extended from the malinobus distance statistic allows for the visit visualization of predicted 95 confidence regions in our regional study this is an effective method for sourcing for sourcing and visualizing many sources in a single two-dimensional plot thank you very much we have time for questions right we do i believe we are yeah we're we're about 10 minutes eight minutes ahead of schedule okay so uh craig is on here so you can craig and i've been beating our head against the small samples for quite a while so i i tried a number of things one is including small samples in the source standard data and that worked not very well uh collimation that was worse and uh i so i started well let me use multivariate statistical analysis the problem with that for archaeology often is you violate the three p times n so you have a lot more variables than cases and so that means you get summer squares uh empty cells in the summer squares across projects product matrix and that means it's going to associate no matter what let's say you have a you can imagine a matrix what they say in hyperspace that's you know 360 degrees it's a ball and in the middle you have a number of empty cells it's going to take a variable over here on the right side the verb on the left side and it's going to stick them together so you always have to be careful of that the other thing going back to my talk or as i was emphasizing barium is that you guys in the great basin uh strontium and barium tend to follow each other in the mineral hosts and so when strontium is low so is barium so i guess you're kind of stuck as far as using those heavier elements cerium sometimes works too but that's still the same problem um what else do i have here i noticed that uh i didn't look at all of it that rubidium and zirconium seem to have a more normal distribution than the other elements uh have you played around with that at all no not specifically but kathy may be able to comment on that i hope so well we usually find that rubidium is among the most distorted because it's furthest from the compton peak right yeah so um that one is pretty problematic i usually find um yttrium is really stable for whatever reason i if anybody knows why well geologically it's because those elements are coming from the mantle and not the crust where you have a lot of variability that's causing a lot of our problems anyway so i think that's part of it yttrium and neobium too but well i guess that makes sense rubidium is as rubidium is farther away it's not really normally distributed it just looks that way so that that adds the issue or another problem for us i know i'm not giving any answers here i'm just i'm just worrying about the same things that you are steve what was the problem with collimation where your count rates too low well uh i don't i'm not sure it's an issue with broker because it's already pretty collimated isn't it i don't know what the size of the aperture is it's not really it's like two by three millimeters yeah yeah so i start at 31 millimeters or 28 somewhere around there and i kept going smaller and smaller so what happens you get attenuation and you start losing energy so it kind of works against you ultimately that's what richard found in that paper that he published and i some weird publication i can remember what journal was yeah we found we found that years ago that with the columnator you just lose all your counts and then your errors are terrible yeah so maybe what luke is talking about is the only way to go i don't know uh refresh my memory uh you were ratioing ppm and not net intensities or counts right uh we're following richard's lead ratioing intensities okay yeah mainly because some of the ratios don't they don't really compute that well when you don't have any strontium yeah and i think it's less an issue of normality too that i'm worrying about yeah because the other thing with the ppm ratios is that it well i mean this is good and bad it adds the errors of your calibration line right right and so in those numbers you have um you also have what your um detection limits are which is you know they will reflect your detection limit like in strontium when it's low um or any other element really right which is good but you don't the resolution you know the separation of your sources isn't as good right well uh i just reviewed a paper for georchaeology from a group in alaska and i don't know if you guys have been talking to them and they've they've been uh using a mahal mahala novus and d system too uh technically i can't talk about it but i'm pretty sure it's going to be published in georchaeology so it looks like that group up in alaska is going the same direction well you know the the jump statistic that you can pull in you know you can pull in all three elements and look at outliers realistically that upper control limit provides a 95 limit it's jump too it doesn't display it in a nice ternary diagram with a whole bunch of other sources so it's doing it's functioning the same thing um but it's hard to do it for you know you know a huge source you know a whole regional study you kind of have to do them individually one at a time and already know that they situates with napo valley for example and then you have to include it in that so it takes a lot more time um than just using the the r program well i've only done that once or twice with jump but at least it identifies where the outliers are so at least you have that information then you can go from there you can then i then i focus on those right i'm out and say what's going on i mean there's other there's there's petrological issues that can occur you get wall rock reaction in a single source and that screws up the chemistry and a lot of times when we get these outliers that's really what it is that do i is there any way to really know that no but uh that that is a possibility right and just fractionation too i mean this stuff's happening in about a thousand degrees c or a little less than that and all kinds of things are happening at least that's what i used to explain it all the way sounds good to me makes sense yeah you analyze enough stuff and you realize that some stuff is just uh these oddities academy and i've seen things too that just respond in a in a weird way it's not predictable um so yeah matt's not here uh right now but he sent me uh a piece and i've never seen anything like it it's not in craig's database it's i don't know who knows what it is but all the rest of them were from serral domedio and there is a mahogany chunk and at serral domedio so it could be that uh when something else was happening in the melt i'm here steve i was just eating just eating yeah you didn't want to see a spit all over the screen yeah yeah yeah has that been resolved since i sent it back to you no in fact uh uh it's it's still sitting here on my desk um i had sent it to mike as well and he couldn't he couldn't he couldn't put it down so did you take a chunk off and do it in any with n a a no no because it's already pretty small yeah i know so no it's smaller than but we don't have isotopic source data so that won't really happen steve where were that one from that's mad i don't i don't know that one is from uh from pickery's pueblo in northwestern new mexico so a total unknown yeah although there is um there is a really glassy red uh rhyolite that that i'm finding in the collections as well um and i wonder if this is because it's so small it's just a fragment off of one of those that maybe got burned or something like that to give it a slightly different luster um yeah i'm i'm going to introduce our next speaker pardon me for for interrupting the next paper is portable x-ray fluorescence for non-destructive pervenian studies in the on sardinian obsidian by valentina mamelia carlo leguli and carla canas exactly okay good afternoon everybody i am valentina mameli a material chemist coming from sardinia so an italian island let me first thank the organizing committee for having given me the opportunity to tell you about our studies our preliminary let's say studies on sardinian obsidians indeed today i'm going to give a talk about the attempt of constructing let me now share the screen before continuing okay hope you are all able to see my screen looks nice it's good okay perfect and so i was saying that today i'm going to give a talk about let's say the attempt of constructing an analytical routine method by exploiting x-ray fluorescence to be more precise an x-ray fluorescence spectrometer that we had in our laboratories for didactical purposes and in order to try to distinguish among the obsidian subsurface of montaerchi last time in during the first edition of the international obsidian conference held in lipari i participated by giving a talk concerning obsidian but from a material chemist point of view so what i am and indeed our work was aimed in deepening the nature of these materials and in particular trying to correlate the magnetic microstructural and morphological properties of this material that can be defined more than a glass in indeed it is actually a nano composite but today beyond this work we actually we actually worked on another matter indeed during my bachelor thesis and master thesis i tried to study sardine obsidian for archiometric purposes also so actually since then i work on other research fields because i work on nanomaterials and magnetic materials so as you can see i'm not working exclusively on this on this field but after this preamble let's start with this talk most of you know knows more than me probably better than me let's say that in the western mediterranean we can identify four main obsidian sources located in four italian islands palmarola lipari pantelleria and sardinia in particular in sardinia there are four obsidian subsources that that are considered as archaeologically significant and they have been named as you can see here sardinia a b2 b1 and c types first of all i would like to highlight that for a provenance study the aim is identifying some qualitative or quantitative chemical or mineral mineralogical variants among the natural sources of a of a raw material that should exceed the those intra-source intra-source variances usually as you know portable x-ray fluorescence is one of the most commonly applied analytical technique for provenance studies indeed it permits qualitative and semi-quantitative analysis and if you have a portable instrument you can also use it directly in in situ and now with this new instrument we can also apply this this technique in a non-destructive manner but of course we know also that this technique has some disadvantages which are intrinsic on the on the physical mechanism so in the relaxation phenomenon of the atom indeed we know that the probability of x-ray fluorescence depends on the atomic number so we are able under air atmosphere to detect elements with the atomic number higher than 17 and so the the discriminative tools are limited to these kind of elements so heavier elements and also quantitative analysis are limited by matrix effects so it means that the intensity of the peaks are not directly correlated with the cannot directly correlated with the concentration and therefore quantitative analysis quantitative data can be extracted only by means of standard of known chemical composition in our preliminary study we have used as I said before an XRF spectrometer that we had in our laboratories for other purposes and it is a LITOS model from Asing SBA so it's an old instrument actually it's I think 20 years old and we applied this the analysis by means of this instrument to different kinds of samples first of all we analyzed mild geological samples because this is somehow an ideal samples in order to see if we could be able to distinguish among the four obsidian subsurances from Monte Archie and then we move on with other geological samples in particular intact samples with different thicknesses and different morphologies and then we analyzed just four archaeological samples so in particular we after collecting the data we analyzed them by using origin software and calculating the integrated areas for decay alpha lines and calculated also the those ratios between these integrated areas and the analysis was done by univariate univariate analysis it means that we checked for discriminative ratios without any multivariate assistance let's say the beginning we tried to find the best experimental conditions in terms of signal over noise ratio because of course if we need to get the integrated area of the peaks we need in particular for those which are not so intense we need to have good quality data and therefore we choose the bigger beam shatter and the measurement time of 16 minutes as you can see it's quite long analysis because this is an old instrument actually so today we have in the market really let's say optimized instrument so we test first as I said before on this method on mill geological samples 16 samples four for each subsource and with our surprise we were able to see that the ssc groups but also the s a groups can be easily distinguished of course in this limited number of samples by this method so to check if this method can work also on wider number of samples and in particular on intact samples with all the the limit of imposed by their morphology we collect from professor carlo luye 38 samples of unknown provenance in order to test it test them in a blind test and also we have the five known five samples of known provenance so it means that these unknown let's call unknown samples were collected from professor carlo luye who did not participate to the chemical analysis and then we checked at the end if our assignment to a subsource of Monte Archie correspond to the actual provenance of the samples and as you can see here we first see in this three-dimensional graph but actually we use different bidimensional and two-dimensional plots in order to assign the provenance to these unknown samples and you can see for example here that a group was assigned to the ssc samples then other were assigned to the s a s b 1 and s b 2 samples so then we checked with professor carlo luye our assignment and we were happy to find that all the ssc samples the s a samples and the s b 1 sample were correctly assigned but unfortunately we had some problem with the s b 2 group in particular because we were not able to assign the number one same sorry one sample the number 515 subgroup an s b subgroup and the other two samples the number 509 and 511 were wrongly assigned to the s b 1 group but indeed they were collected from the s b 2 outcrop so then we test again the method on just four archaeological sample and in this case the assignment was compared with previously with the previous sorry particles induced x-ray emission assignment and we had a perfectly matched but as you can see we have only four samples so to sum up we were able with this method in this blind test to assign correctly the 92 percent of of the samples with an eight percent of error correlated associated with the s b sub-discrimination but of course this is for us just a starting point because we are aware that we need to have a statistic significance we have to enlarge the number of samples to be tested and also I think that it should be simplified the data analysis step and for this reason I'm now working on trying to to find other ways let's say for example changing the the analysis from discrete variables as the intensity ratios to continuous variables so working directly on the XRF spectra in order to avoid any any loss of of information and also we I'm also working changing from univariate to multivariate analysis because of course we know that for a provenance study we need that this this should be effective of course but it should be also routine so it means that we should apply it on a lot of samples so for this reason now I'm working on a different aspect also we recently acquired a new instrument the tracer 5i from brooker for another for another project but we have acquired also the matrix matched calibration developed from brooker in collaboration with the with the group from professor glas cook and and so we are going to compare our data with those obtained on directly on concentration by this by this instrument and also I'm going to work on on changing the for example the data analysis based on the intensity ratios moving from the origin software to other software for example pi mca that is specifically developed for x-ray based data let's say and so with this I come to the end of this talk I would like to thank professor Carlo Luye professor Carla canas for their collaboration and also European Union because they are finding funding my fixed term research position at the University of Calary and all of you for your kind of attention thank you very much we have we have some time for questions I believe a few minutes for questions well I guess I have a comment so for XRF data in the discrimination of sardinian subsources I mean a bivariate plot of rubidium and niobium and strontium to niobium has fairly good discrimination between them yes indeed we have different elements that can be used to discriminate in particular for the SC group which is the easiest group to be distinguished from the others we have a higher I can show you maybe if I can again share my screen I can show you please do okay I can show you some data if I be able to move on on the presentation let me check if I can do it okay so for example here I've just plotted the 16 XRF spectra for the milled samples okay and I've just grouped them based on their provenance and as you can see here we have for example for the SC group we have a higher intensity of titanium of course here the spectra are normalized for the highest peak that is the iron one so the most intense peak and so it's like seeing all these peaks as over the intensity of the iron okay and you can see here for example that for titanium we have the highest content for say the highest intensity for the SC group and the lowest intensity for the SA and but of course the most interesting peaks are the rubidium and strontium ones but also zinc for example because we have a higher intensity for the SA group okay also for the manganese as you can see here so I think you can you should adapt the let's say I have adapted the the method to what I was able to see in my XRF spectra so it depends on the instrument that you have let's say so in my case the goal was to adapt the instrument that we had for this purpose so we did not acquire an instrument only to to work on this issue so I just see what information I can extract from this analysis and try to use it for this archaeometric purpose but of course this is just the beginning thank you Valentina right with that I will introduce our next paper Obsidian Hydration Dating by infrared transmissions spectrography by Christopher Stevenson Egan Lydford Alex Jorkensen and Alexander Rogers Thanks Lucas welcome everybody good afternoon let me screw the chair we're good to go yeah it's great okay this project is part of is the methods development methods development section of a larger project concerned with looking at Maori social networks in New Zealand and their development over time and XRF is a large component of that but my responsibility is to add the chronology to these various archaeological sites of course in conjunction with radiocarbon data which is because of late time the late time period in Maori society is often problematic um so this this is this talk reflects kind of experimental from experimental results dealing with and using infrared spectrography my my role here is to provide the chronology for these archaeological sites and by using city hydration dating which has been tried two or three times in New Zealand starting very early in the 60s um and you know it's gotten better and better over time but no one's ever been satisfied so we're taking one more stab at it and the New Zealand crowd the hard crowd and they really have to be convinced so it's not really the show me state of the of the international world and so you know of course the problem is it's been the same for decades it's converting the amount of gain in diffused water into the surface of glass into time based upon an experimental rate and so it's the question the focus of this is is coming up with a new methodology to improve our precision um and it's really it's very important in New Zealand because the the chronology of that um that culture the Maori culture is very short 80 13 50 to 18 50 approximately 500 years and so the resolution becomes a big issue and in addition in addition to the great development um and this project seeks to kind of improve both so uh let me start out with with kind of I want to suggest a change in terminology that that is important I think because it redirects the way we think so traditionally we talk about the hydration layer on the fitness of the hydration layer and those are set up concepts that I have what they do is they lock you into a mentality or way of thinking and it's very hard to get out of it because you start going down an intellectual path uh looking for some inappropriate solutions because the hydration layer in fact not a layer okay doesn't meet the definition of a layer it looks like a layer and yeah it's when it's small and and the 600x magnification it looks like something that's resting on the left but in uh if you look at much thicker hydration layers like the one from Africa here it's 30 microns thick uh and you have to see that actually it's not a layer it's no sharp discontinuity um it's actually a gradient and so uh and that's and our our kind of recent well our conceptualization the problem is that it's of course it's a diffusion process and there's no layer development there's mass gain there's molecular increase of the number of molecules origin might be in this case an absorbance increase when you use infrared spectroscopy and so that kind of aligns our terminology with with the problem and so if we um but it's like I like quitting smoking you know kind of hooked on the terminology so if I'm making the state you know uh forgive me so we have our our mass our our uh diffusion zone called a diffusion zone as well zone of diffusion zone of accumulation and the conceptually uh it's going straight forward water on water vapor it's absorbed onto the surface of a newly napped artifact and over time it fuses into the into the surface of the glass to form this is water concentration gradient and if you look right below it if you look at the the dark blue line here represents the the the diffusion uh the fusion profile this is done by secondary ion mass spectroscopy sims so you can uh this sets up across defines the problem nicely there's really no uh there's really no depletion um of the of the obsidian and water is the kind of mobile species that works its way into the glass there and there's you know there's a hint of some kind of enrichment surface but basically once you get past that that first type of micron molecular molecular water is the only active species um that's creating the um the stressed zone of of molecular water so okay we look back and look at this at this process in detail and over the last year or so just kind of get our thinking straight and and to define what the mechanism is and uh this work that based on kind of I knew it's an early newspaper by uh followed by Korota and his research team out of japan who uh who thought through the problem very well and in when when glass is what when glass is molten uh there is there's the first stage of the fusion that goes on and that offers the structure of the glass for the water it's in the magma in the magma chamber or in the vent uh is transformed into hydroxyls and hydrogen and what it does is it connects with the glass network and and it opens it up so if you see here you can see that hydroxyls link themselves to to silica and hydrogens um link themselves to um to uh the oxygen so what that does is that put doorways or pathways establishes pathways that are that run throughout the glass so there's lots of void space there's lots of space in that glass and molecular uh and once the glass cools and becomes solid those pathways are still there and then water that comes in from the surface navigates through those pathways so in the first part water breaks down into into uh converts from H2O to hydroxyl but once once the glass is established water diffusion consists only of um left with water diffusion that inside by uh that inside by by Korota and his team confirmed the correlation the correlation that we've been operating on for many years and that is that the hydration rate for molecular surface molecular water is is a consequence or varies with the number of pathways that are in the glass so so that that that's the mechanism and um in terms of look uh establishing methodology for for obsidian hydration data we search for solutions that rely on that that are appropriate for the for the measurement of this state for a mechanism so uh we we we have in the past we have used infrared transmission i i've used photocoustic spectroscopy for a number of years to measure water on the surface of glass and uh it was okay but it didn't have as you'll see didn't have the range that we need range in the mid infrared to monitor all the different expressions of structural water and issues the water that's on the surface of the glass so we discarded that and we want to transmission spectroscopy and um it's it's suitable well it's suitable because it can excite and measure uh the vibrational responses of water when they are subject to infrared light so in the graphic here uh in this approach we simply put a sample a flake or a cut section from a flake in the path of the infrared beam and the water molecules will vibrate absorb energy and give us an infrared spectrum and so the uh and the water will express itself at various wavelengths which i'll show you so um and infrared mid infrared is very sensitive to measuring water in glass normally an infrared you know water is an interference but in our case it's very strong interference but in our case it's um it's the amount of water or the number of molecules of water that we actually want to be measuring so in our in our application in New Zealand New Zealand is is a it's not a transparent glass and so it poses a little bit analytical difficulty but uh if your glass is transparent like i've been seeing today like for the chikniata source or the pachuka sources the really good news is that this this technique has potential to be non-destructive to the artifact um and give you chronometric information so that that good news is coming up at the very end um but i'm going to show you the methodology that we have here for non-transparent glasses so we can break our analysis so we have an artifact with a hydration zone on it uh of accumulated water and we want to measure the amount of water it's accumulated in prehistory and then we want to measure the amount of structural water um uh then glass you know once it's solidified once it just came right out of the volcano because it's it's the amount of structural water that determines how fast the surface water can penetrate into the cdm so there's two analytical steps uh and the first one is to measure the structural water of the of the sample and aren't prepared on a pair of cutting thick section or cut cut and polished coupon and it's straightforward using using beer Lambert laws they're around for a century um the equation here is that uh is that on the top here no concentration of water weight percent is 18.3 which is molecular weight of water infrared absorbance thickness density and then the more more absorptivity and that that being that being how much energy is absorbed um by water species it's just a way of life so um so in applications wise um more absorptivity is already figured out back in 1984 by luminators and uh that's well established and then uh so we're interested in thickness of the sample not of the layer thickness of the sample and thickness of and the glass density two variables necessary to establish concentration and so we have two infrared peaks the the 4500 peak is the one that we use to uh measure measure hydroxyl or structural water and we measure here's a here's a polished coupon and we simply measure it with micrometer and then we measure its density by the Archimedes method right by the immersion method you can immerse it in water or you can immerse it in toxic heavy liquid I prefer the heavy liquid um just because it involves air bubbles and and then once you pull in once you once you pull in those two parameters then you're able to calculate then you have all the then you have all the information needed for the beer's law formula and you can calculate um structural water concentration in weight percent and for the most meant for most of the recipients and we want to deal with we want a very low structural water we want something in the zero to point one rate zero zero point one range so that's the first analytical step step count uh calculate the structural water which will allow you to predict the hydration rate then the second part is is that we have to actually cut a little slab off off the off the flake to make it transparent it has to be thinned down to about under a millimeter and then we simply pass the infrared beam through it and it gives us our our um our structural water 4500 okay and it gives us this what we call combination peak of OH and H2O at 3750 so to get mass gain we need uh mass gain is contained within this peak right it's contained within this peak here oops sorry mass gain and hydroxyl arcane are this this peak is like a summary statistic so what we want to do is convert the structural water into the absorbance here uh using the extinction coefficients for the more absorptivities and then because this peak will come in just under the big under the 3570 peak so it's always going to be less and so the difference is the mass gain that's occurred over time so this little trick which is the innovation here is that we we want we convert we convert 4500 you know the absorbance of 4500 the absorbance of 3570 and subtract the two and we're able to do that because we know the more the more actual activities you know this ratio between the two we apply that ratio to the absorbance at 4500 we get a conversion so so uh so you get a mass gain so you have two we have two critical variables that we need to compute a rate a date we have a mass gain uh molecular water and we have a structural water concentration that allows us lots to predict the velocity but what about error so we looked at we looked at our errors uh coefficient variation because we have four potential potential contributors of error the absorbance density thickness and then extinction coefficient and we looked at we looked at those individually and and there's some and the sum there and we come in at an error of about 1.3 percent for those four different four different measurements net net so this this error applies to calculation of structural water this this error applies to the calculation of of gain molecular water gain or hydration gain so because you're measuring two two peaks on the same sample um density and thickness cancel out of the equation um but their contribution to the equation to the error is almost infinitesimal so you basically end up with the same error just under two percent so that that's encouraging so we have we have the variables that we need to calculate our rates um we have uh we have our our structural water we have our mass gain and of course hydration the fusion processes are temperature dependent and so the hydration temperature is estimated soil growth of humidity is estimated and then uh below this are our our our calibrations that that relate structural water to hydration rate so we get for the arenas equation that allows us to implement to extrapolate from our high temperature experiments at 150 degrees um we developed prediction equations for for the activation energy and for the pre exponential and so this is the pre exponential this is activation energy and that is used to calculate the hydration rate in absorbance for your archaeological article so um so it's all kind of cut and dry right now so it's the methodology set up these are easy equations um they they look coherent but we're still in the evaluations processes though how widely applicable they are so this is so this is the method for um for opaque insidians uh but but it does it does have its problems right it's destructive um uh and samples aren't transparent so you know a lot of insidians you know have to be um you know samples have to be thin and then the biggest drawback is in this is in this uh conversion process with a small amount of error for the 4500 measurement for the 4500 peak becomes amplified because you're actually multiplying it 125 so uh so you end up in the four to five percent error range so that's not that that's not good um it's not tedious but it's not good and so there's room for improvement there um so the our next step um this has been to say what can we do for transparency so looking at all all the insidians in mexico in the boxes this has some uh applicability there and the there's a there's another water band out here for 5200 that we didn't look at in this first application because uh because in the mid-air for it um you know you have a your water diffusion zone it might be only one micron i shouldn't say that i can't say it there uh it might be it's going to be you're going to have very very low molecular molecular concentration on the surface and so that 5200 band is is not very sensitive in the mid infrared but that's because the mid infrared loses energy in this region but if we shift to we just simply change the source in the same instrument uh to a flight light source our our energy our energy our background energy um becomes very high in this region now and so we can measure the 4500 peak you see here and we can measure the 5200 4500 peak is your structural water and then this 5200 peak is your surface water so uh we get both variables that we need without any any subtraction and without any um uh conversion um and and this now becomes detectable uh non-destructive under and so but how precise is it well don't quite know yet these results are only about three days old but one of our samples is also con that has an optically measured hydration layer of two microns um and um what this green peak represents is actually four microns because the infrared beam passes through two sides of the artifact so it picks up two surfaces so it doubles the response so uh you simply have to divide the the absorbance in half which you can see here for like the uh the red the red um the red spectrum is the shulveria source in georgia and um this is from shulveria shulveria gora um it's a site about 5000 years old um and it has a well developed peak as well so um um this if you have a transparent obsidian um the analysis of uh degree of hydration comes substantially easier and um and and straightforward and it becomes non-destructive um then what reason becomes non-destructive you do have um but we still have to calculate concentration we still need a sample thickness in the way we the way we get around this irregularity of artifacts is that we aperture down to a very very small area and so essentially there's no lateral variation in the thickness anything is one millimeter aperture so that allows us to to calculate beer's lock for our structural water so that's a lot to absorb in 15 minutes but uh i think the hydration dating had now had a new approach uh it's it's consistent with the theory and um you want to try it out and let me know thanks very much thank you christopher yeah that's a lot that's a lot to take in that's a lot to take in actually it's very simple it's actually two analytical steps um for for opaque obsidians and it's just one analytical step uh transparent for the for the what i just showed you on the for the transparent material and you put your sample within the beam path of the spectrometer 30 seconds later your analysis is done so it's it's gotten substantially easier over the years any any question can i ask you yes please uh with a huge respect christopher thank you for the interesting presentation my question may sound a little bit stupid but i want to ask you so uh how hard it will be to use the method you described for the sources which has very extreme environment i mean the for example very high percentage of precipitation and so on how it will affect on your methods thank you precipitation um well the most artifacts will be in the ground and um you'll have you'll have the rainfall you can have white precipitation but as long as the soil is maintained at 98 percent humidity you're fine okay the the rainbow the rain is not going to road the artifact it's only if the if your soil ph is very outgoing and you have heavy rain that will start to dissolve the artifact you know with exposure over the centuries thank you yeah that's been very noticeable in places like you know central america and so on who i think i went over i'm sorry and no problem so we're in the midst of a little coffee break here we'll resume the the next paper in about 10 minutes so feel free to take a little short break or if you want time for some extra questions some of us are here that have talked earlier in the day i was really excited about this approach because you know for years you know i've been cutting and grinding up obsidian and uh stringing the porous and tedious correct isn't that right correct and you know in my dissertation i think i did 900 slides you know grinding each one down i just said i'm not gonna do this anymore and it took a while to develop a method like this that the one of the impetus impetus one of the imps this is for doing this is that you know the much harder to do obsidian dating destructible you know as you as chariots put new rules on it and indigenous groups don't want their your artifacts destroyed and so that the directive is to come up with a non-destructive method and maybe in the other part of the world it's not an issue yet but i'm sure it will be probably will be and sometime in the future so uh it's kind of it's kind of it's kind of merges with the upcoming policy of archaeological analysis so that's kind of good thing yeah so um but you know i'm always you know it's hard to come up with really good test cases because of the difficulty just finding stratified deposits just finding you know their obsidians and just finding cities that haven't been you know altered by post-death positional operation and then the issue that hasn't been discussed here is the it's the actual estimated thermal history and i know it's Steve Steve brandt is still here um you know working in in africa and on these very old sites you know i'm here hey there it is uh working in these very old sites coming up with a long-term estimate of temperature is is quite not very difficult we have a solution to that i think and that is why looking at the refractive index of the surface diffusion zone i'm just writing a chat message to you about the oh okay you can use it so we'll talk some more so uh but but it required but as it's done today it requires temperature monitoring of your archaeological site it just takes a year so uh you know signing your research is on you have to think a year ahead so you gotta put sensors in the ground for a year um that's it gets it's part way and you never know about the thermal history of the artifact you know it gets buried it's picked up and recycled you know it's transported hundreds of miles away those from the mount and down below so um that's that's a challenge but in areas uh for short short for short free histories you know it can work out pretty well um for very long free histories it's it's it's a challenge Bruce how do you think it compares cost-wise is it is it that's the other driver you know you can you can do great things with half a million dollar instrument like a synth machine right but it has to be it's usable so the infrared is uh basic instruments around 28,000 right that's that's really all you need um in this in this uh the results I've shown you here so you need a basic fTII you don't need any fancy options on it you know a white light source um and so that so that's like a good microscope is like $15,000 to do it the traditional way you know for $2 more you know you can you can have a new technology but you know the spectrum is our name it's they're not being in the street it's five years you have to buy a laser it's kind of like those xrf machines they break down um so that's why pep talk for obsidian dating I hope somebody uh takes me up on the offer if you have a clear obsidian you have dark obsidian don't talk back from our our short break our next paper is titled archaeological significances and geochemical characterization of obsidian sources in central highlands central japan by wavelength dispersed xrf and la icp ms by yashimitsu suda okay thank you for your introduction so let me share my desktop okay can you see my slide okay yes that's great okay all right let me start my presentation so it's great honor to be able to be join this visual meeting of international obsidian conference in baku so today I'd like to talk about archaeological significances and geochemical characterization of obsidian sources in the central highland central japan it is well known that there are five big or major obsidian sources around kanto plane which are called central highlands takahara yama hakone amagi and kozushima shimada etou indicates that the obsidian from central highland was most abundantly used as a lithic raw material in the kanto plane during apocalyptic also to look at the change in the use of obsidian in the kanto plane there are two big change in the use of obsidian from central highlands as the periods of the beginning and ends of late apocalyptic many researchers have been pointed out the relationship with the impact of lgm but the discussion based on the evidences of paleo climate change and human activities in this source area is quite insufficient so based on this the group of major university have been performed the archaeological excavation and paleo environmental research in helopolisites and bog at central highlands and yoshida etou reveals that during late apocalyptic the landscape of helopolisites and bog was characterized by alpine vegetation with rocky stretches and the tree line is located just around the level of helopolisites and bog after that the landscape was changed to be forested due to warming and rising of tree lines also more than 2000 of obsidian artifacts from early apocalyptic to german periods were excavated from this helopolisite we had already performed the archaeological analysis for these obsidian artifacts but there are still many difficulties on the provenance analysis of these obsidians so the aim of this study is to establish the methodology of provenance analysis of obsidian in kirigamine where the helopolisite bog and the site are located also to reveal the human activities concerning the procurement and consumption of obsidian in kirigamine source we try to improve the methodology of provenance analysis of obsidian artifacts to identify the procurement area and the point and this kirigamine source what you can see here is the locality of obsidian sources compiled by previous works the previous work has already already indicated that there are 21 points of obsidian sources in this kirigamine area and the obsidian in this area can be classified into these 11 chemical groups also looking at the locality of obsidian sources in this area it looks that there are no clear relationship between the locality of obsidian sources and their chemical composition so this indicates that to improve the provenance analysis of obsidian in this area we need to make clear the relationship between the locality of obsidian sources and their chemical composition for this we again performed the field survey and field sample collection in this area and 109 samples of geological obsidians were analyzed by WDXRF and laser ablation ICP-MS you can see here is the instruments for chemical analysis of obsidian used in this study all samples for analysis were used into such this gas speed at Nagasaki University and using this gas speed we performed WDXRF analysis at Meiji University and laser ablation ICP-MS analysis at Kyushu University the compositional data by this study and previous studies had indicated that the obsidian from these major sources and counter plane can be discriminated based on these variation diagrams okay now let me explain why the magma of different composition had been formed the first process first reason is the process of magma generation in other words the composition of magma source is different the second reason is the degree of magmatic evolution is different even though the magma source is the same let me explain further explain what is the magmatic evolution beneath the volcano magma is always gradually crystallized in the magma chamber during cooling so because of this process the composition of magma are also gradually changing we call this process magmatic evolution also in such variation diagram composition of variation diagram the first process is shown by such these compositional clusters for each magma source on the other hand the second process is shown by such this trend line for each magma source so namely chemical groups of obsidian can be classified according to the difference of magma source and degree of magmatic evolution you can see here is a compositional variation diagrams of obsidian in Kirigamine from this diagram we can identify five compositional cluster and one trend line these diagrams show the composition of obsidian in the trend line from these diagrams the obsidian in the trend line can be further classified into more eight chemical groups based on the degree of magmatic evolution so this study indicates that the obsidian in this area can be classified into these 13 chemical groups however looking at the locality of obsidian sources in this area the relationship between the locality of obsidian source and the chemical composition are still unclear so to perform the provenance analysis of obsidian in this area we need to recompile these chemical groups based on their localities let me show again the chemical groups of obsidian in Kirigamine the group forming the clusters can be classified into these five groups on the other hand group forming a trend line can be classified into these eight groups but taking into account also their localities these are recompiled into most evolved intermediate and less evolved ones you can see here is the locality of obsidian sources for each recompiled chemical groups as you can see here to perform this recompilation for these chemical groups the relationship between the locality of obsidian source and the chemical composition become rather clear based on this we perform the provenance analysis of 40 obsidian artifacts excavated from these helopora sites you can see here is the discrimination diagrams based on wdxrf analysis in this diagram the composition of obsidian artifacts are also plotted the result indicates that roughly two-thirds of obsidian artifacts were classified into these mt chemical groups and including some w, k, a, g, h, o1 chemical groups this result also supported by these discrimination diagrams based on laser ablation icp ms analysis so this study indicates that the prehistoric human center and this helopora site exploded obsidian at around the area close to this helopora site of course this study need more investigation and need more provenance analysis of obsidian artifact but this study was able to be shown the relationship, correlation between the locality of obsidian source and the chemical composition finally let me talk about our ongoing studies for the provenance analysis you can see here is the instrument installed in our laboratory of course these instruments have both advantage and disadvantage on the analysis of obsidian artifact but based on this we are going to establish such this methodology of provenance analysis of obsidian artifact flowing this flowchart also recently from presented such this open source collection of reference samples of obsidians we want to follow up this work also in our laboratory we are making such this a package for the collection of obsidian reference samples and nearly open the website of obsidian source online database we hope these become valuable resources to promote the provenance study of obsidian artifact in archaeology thank you for your attention that's all thank you yashimi too any questions hi hi yes a really interesting presentation and I observed it in a difference obsidian source that that the obsidian flows was the last eruption or the last volcanic event mm-hmm in the evolution of the magmatic chamber some process produce the obsidian at the last volcanic events yeah I can it's now I cannot imagine the volcanic activity itself in this area but so some of age data is obtained in this area but it's quite narrow at a range of a few million years so but so I these data compositional data indicates some different composition of magma flow or magmatic activity occurred in this area and so to look at this distribution I suppose the last eruption of of liotic magma produced such this obsidian is such this green green one because it overlaps the different another another kind of obsidian dissolution area but so it's still become it needs more investigation for the linkage of the geological interpretation it's okay yeah okay I believe your presentation are really important because you can define the chemical range the chemical change yeah and that means if it's really important to study the geological evolution of the volcanic process yeah yeah exactly and in in Mexico many obsidian source are around big volcanic mm-hmm calderas uh-huh yes what is the composition of variation yes and the the volcanic process start of a salty composition salty huh and after when the caldera a big explosion of the caldera the caldera fell down mm-hmm yeah yeah start uh radial fractures uh-huh fractures at these fractures after start to grow the obsidian domes uh-huh uh-huh yes yeah the chemical study mm-hmm identify different source difference uh-huh okay but uh according with your presentation is the same volcanic origin this the same magmatic evolution mm-hmm this is important because sometimes uh the chemical study said this obsidian is from here this obsidian is from here and is the same area yeah yeah exactly yeah yeah I convinced that so the composition of magma is completely represented in the composition of obsidian yeah it's very important to to the composition of the obsidian to reconstruct the volcanic history such like that I think so that so also I want I didn't mention that so this this source including some of the uh deposition derived by the derived derived from phylocraft stick from phylocraft stick also including including in this obsidian source such as takayama source is a is a deposition of deposition originated from the phylocraft stick from but so volcanic event is located here higashimo says I said that higashimochia so it must be uh volcanic event occur here and the volcanic phylocraft stick flow is yeah is flooded it's it's many around the area but so I uh detail of the history I cannot now reconstruct it I would also suggest that that applying you know perhaps dating techniques you know fission track dating or whatever might help untangle uh you know these complexities because you know these aren't unique uh to to just Japan you know there's all all around the world people are facing these issues and defining things yeah I think so that so we want to so perform such this so uniformed age dating analysis such as uh potassium argon dating of glass glass parts but so we need to uh uniform the method of the age dating method I uniform the method of age dating to compare the such this frequency frequency of the volcanic event I think so thank you thank you yes yes you should meet two uh very good presentation wonderful questions very very engaging topic okay thank you for your comment all right our next presentation is an inter instrument calibration and data comparison for xrf analysis of obsidian by robert tycott robert I hand it over to you thank you you can see me and hear me yes sir okay well I'm not sure that I'm talking to the right audience here about this because we pretty much know what we're doing but keep this in mind for all of your colleagues people you may share your instruments with uh and that kind of thing okay I'm going to talk some more here about the use of a portable xrf how we calibrate the data that we produce focusing specifically in the mediterranean area dealing with archaeological and geological data and comparing this a bit to analyses that have been done by other methods as you know there are both advantages and limitations to using xrf instruments I'm not going to read through this whole list here but in particular the potential for doing non-destructive analysis on artifacts that's become a very big issue these days on archaeological samples the ability to transport the instrument to doing work in the field or museums especially in foreign countries how quick you can do it what it costs and those kinds of things those are the pluses involved here but we always have to keep in mind the limitations we're not putting out of business doing neutron activation and icp and other kinds of analytical methods here in particular we're only getting results on the surface of the sample so if we're being non-destructive that's not the same thing as what we get with those other analytical methods in the case of obsidian we usually don't have too much problems with weathering or contamination you know we're not doing pottery or something here but nevertheless that's just something to keep in mind and in particular again where we're analyzing original samples heterogeneity while we call obsidian glass it's not quite the same thing there can be in different sources of obsidian different kinds of crystals and so on accumulated in there and so where we're going and analyzing what spot we're doing that can have a bias in there as well and then we've heard today in some of the recent presentations issues where we're dealing with how thin the sample is or is it so small it's not even covering the area that the x-rays are coming out of the instrument we have to deal with that as well okay time to move on from this we already have seen where the different sources are in Europe the four islands that belong to Italy two for Greece and then parts up in Hungary Slovenia and that other country up there Ukraine right okay and we know what obsidian looks like how you have cores producing various kinds of tools and we have different kinds of research questions depending on our particular projects where the particular cores are being produced compared with the final products and so on I'm not getting into that here today but ultimately we want to go and match where the artifacts are found compared with their geological source and deal with what kind of distance trade routes etc etc how they change over time. In some cases it's possible simply from visual examination to distinguish some sources but I'll just say right off the bat in the Mediterranean and in the area in particular that I'm working on today that's not sufficient by itself we have to do some kind of chemical analysis and in particular when we're dealing with subsources and not just the general difference for example between liperi and pantilleria. I also don't really need to show you the periodic table here but again to emphasize the difference between portable xrf analyses and what is being done by ina or icp for the most part we're focusing on five trace elements maybe adding the iron and a couple of other elements in as well and so we just have to keep that in mind again also for various issues concerning the settings that we have for the pxrfs in particular dealing with lower elements where they are getting the x-rays get absorbed in the air but that's not the situation for these five trace elements without even doing any kind of calibration if you want to just go and compare results that you get from geological samples with your artifact analyses just looking at the peaks you can go and see how good an overlap in all that we have here for the iron and those five trace elements over here clearly coming from the same place whereas in the case in the lower graph clearly coming from different sources with the red having less strontium but more rubidium and zirconium and other elements here and so this brings up a little bit ahead here is the whole issue of how do we compare one set of data done with another set of data and I can just say right off the bat I've been using a pxrf for 15 years but not with the same pxrf instrument I have gone through more than a half dozen different ones and so I've beaten around the bush I guess by reanalyzing the geological samples that I have with each one so I don't have to worry about the calibration in that sense but I'll come back to that again a bit more here anyway the just showing the latest instrument that I have and I know some others again have these two okay I'm specifically focusing here on the research that I'm doing in the central Mediterranean and particular in the area within the green ellipse here where there's a whole bunch of potential geological sources being represented coming all the way from sardinia and leapery and palmarola in the Turinian coming all the way from Melos off the map here to the east and also from the Carpathians up here in the northeast and in addition and I've shown this in some other work and we'll show once again how important it is to distinguish the subsources within each of these islands or other source areas and this I discovered back in the 1990s looking just at this one site as my first investigation on sardinia where looking at the difference between the sources a b and c how they changed in their usage over time from the early Neolithic to the later Neolithic with the sardinian c subsource becoming the dominant rather than a minor portion okay and so this is why it's really important to assign artifacts to subsources and deal with the modest chemical differences between subsources versus comparing simply sardinia with leapery and other central Mediterranean groups okay I forgot what I was going to say about this but just showing it basically with pxrf and just using a few of those trace elements rubidium niobium and strontium how we can go and distinguish palmarola from leapery from sardinia pantalaria milose and and carpathians you'll also notice why as some questions why is it such a broad variation for milose well that's because we're doing a ratio here and the amount of niobium is very low so simply the numeric statistical part there is why that is so widespread but in any case fairly straightforward to distinguishing between the different islands but yes here in the carpathian area where there's many source groups and subgroups and I thank katalin for organizing the team meeting there and the opportunity to collect a bunch of geological samples showing that again just using a few of the trace elements we can go and distinguish things into separate groups but also notice some things which do not separate as easily the 1a versus 1b and and so on okay coming back to my area here on leapery we have two major or not necessarily not all considered major but two distinguishable subsources the major one in gabalato gorge and the smaller one in canettodentro here in the southern area and uh you know what artifacts look like and the zillion artifacts to be analyzed anyway we can go and distinguish not only those two groups canettodentro and gabalato gorge but also the one other ancient source montadela guardia which did not produce pieces of significant size and therefore no artifacts have been found yet made of montadela guardia and i'm not showing on the map here the more historic eruptions but those can also be distinguished as well in case you didn't have good context for your artifacts and we can also distinguish the four ancient subgroups using laser laser ablation icpms and yes we can do it with the portable version of the brooker as well using those same few trace elements okay another example is with milos in greece uh where on the same island there are multiple uh subgroups the deminigaki and the stanikia or ademas subgroups again using just a few trace elements well let's look at some numbers uh now here or i'll talk about the numbers you don't have to read them all but one of the early studies i did was simply comparing what kind of variation do we get in the numbers that is the concentrations when we're calibrating them but the analysis was done with different microamps whether we used the filter or not and whether we used a vacuum or not well this depends on just say right again in the case of the vacuum that has no effect on the elements that i'm showing you here but it certainly has a big effect on other elements that are used in many cases for determining the composition of obsidian so just keep that in mind it really doesn't make any difference what the microamp amperage is and we have a couple of differences here in iron and some others that are bolded but that in itself is not really an issue but do keep in mind that even though this were standard pieces of obsidian and it was the same standard run again obsidian is not totally homogenous so putting it back and forth on top of the pxrf that accounts for some of the variation that we're seeing here as well okay so as i mentioned over time i've had a number of different versions of the bruce brucher tracer three series and two versions of the tracer five i'm not going to read all the details here but let's just say again that there are different fundamental aspects between these two general models in terms of the detector there's even a difference in the composition of the filter that is used in both of these cases and how when we're dealing with our ultimate numeric calibrated data where is that actually coming from well in the earlier brucher tracer three versions this was coming from brucher which came independently or whatever from analyses done in missouri by my glass cock and colleagues um but it was in it came with the same set of software for running the calibration they were not doing this on my specific model or that is on the specific physical one xrf unit that i purchased okay these were done like on the whole batch in general so that could be variation simply from the manufacturer and the and the pieces the detectors and so on in each of these instruments in the case of the version five though i have gone and analyzed the 40 geological samples from my glass cock directly myself and put that into the calibration program so it doesn't matter to me what kind of variation there is between mine and some of the other people who have these so keep that in mind here when we're dealing with numbers okay so i'm comparing here specifically for the gabalato gorge subsource on leapery with the various instrumentation here the three dash five three a dash sd the five i the five g and also laser blation and in a data as well and you can see in the case of the rubidium that we have lower values for the rubidium and the yttrium higher values for the zirconium with the tracer three sd for some reason the iron seems to be much higher with the five g anyway there's a lot of variation here and so it's really very important to keep this in mind when you want to compare your artifact analyses with your geological samples or that is you need to go and keep doing your geological samples for a direct comparison okay looking here on milos and the two different subsources there deminigaki and stanitia and again looking at the variation between the different instruments of my tracer three and five and also comparing to marina millich's publication where she's using an olympus inovacs and again you can see that for some elements in the green you have lower lesser values some that are higher in the red and so everybody has to just be very specific and careful with what they're doing okay here is millich's graph showing the comparison that she made as well between her pxrf and icp and pixie and regular xrf acknowledging that there are differences between different instruments as well as between different versions of the pxrf which is what i'm showing you here when you're only dealing with two sources that do not have similar values well we don't have to really worry about it it's because everything is still in the same broader group concentration areas but i'm showing you here the differences between again big difference between lipery and milos we don't have to worry about that but let's look here within the milos what's actually going on between those two different subsources you can see the variation that we have just within lipery gabalato and the same thing is true on the milos as well in particular the deminigaki group which i have here now in red that the different methods the different instruments that were used those offsets make them overlap quite a bit with the stanikia values so therefore you can't mix and match results that you're getting from one version of the pxrf with another you have to go and analyze geological samples and with the same instruments so when we analyze artifacts as well as geological samples from potential sources in our regions using the same xrf calibration is not an issue but changes over time in xrf models and detectors and keep in mind they do not last forever they do not last as cars do in many cases so just keep that in mind analysis of lower z elements and in the calibration programs require great care and direct numeric or graphical comparisons of mixed data always important to have enough geological samples also to show the range of elemental variation which may reflect multiple eruptions over time and space well most of us have our own geological set of samples from our study region beware of undocumented subsources in some cases as well as potential imports from sources not in our collections review carefully studies by non obsidian specialists publications and things i won't mention some that i've reviewed who have used pxrf but without geological samples for direct comparison thank you pxrf is portable precise rapid and less expensive but requires the use by somebody trained in x-ray analysis and data processing thank you rob questions for for dr tycott okay may i have a question one question so i want to know about your opinion that the data obtained from the pxrf analysis it's shareable for among the researcher and the robot laboratory i cannot so please please let me let me know your opinion or or opinion based on your data is it okay sure okay yeah so you can you understand why question okay i'm sorry i didn't understand what you just okay said okay shortly speaking so the data obtained from the pxrf instrument is shareable or shareable among researchers or it is difficult to it it it's still be difficult to among the researcher to share the data or pxrf analysis it's really no different than comparing results that you get on your ic pms machine and what my glass cock produces from mar or on your pxrf there's variation just based on the calibration software that's being used and and various things like that that you have to be careful of okay so my so basically so in my opinion the ic pms data and wdx data is so quantity data is quantity data very precise precise precise data so it's shareable and among researchers but so pxr is quite i think that to call it's quite too difficult to it depends on the machine itself the data so i i'm so asking about this i think that the pxrf is as quantitative as any of the other instruments the precision uh that we get on the analyses is just great it's the same issue with all instruments uh is simply being able to compare one with the other okay okay i can us on to thank you thank you i have a question uh lucas i can uh what about using the microprope and comparing px pxrf data with a microprope has anyone ever tried to do that uh yes i actually use the microprope for my dissertation work uh long long time ago uh and focusing of course on the major and minor elements that are detectable there but it's the same issue it and we can go and analyze many of those same elements with the pxrf using different settings and vacuum or helium flow but you just can't go and make a graph using results from both different methods without being able to analyze some of the same objects and show what kind of offsets there might be so that that highlights the importance of standards in some ways absolutely may segue into the next talk may ask one more question sure i don't know if there's time also to rub uh you know in our Ethiopian research you use the same instrument uh to look at uh the gamma obsidian but you never publish the details and uh you know we've asked you a few times for the details can you ever get us the details because we can't compare it because it's the same instrument uh we haven't published that yet uh because we haven't identified the actual source of locations and that is in progress uh and well it's been delayed being able to go back there uh again because of the COVID but we do have we have source information uh but you did publish it you published minor information not enough really to do any comparative work with not yet i see okay thank you rob yep our last paper of the day is an SEM based micro XRF and portable XRF spectrometry study of Mesoamerican obsidian tablets by Edward Vicenzi, Meredith Sharpe-Noyes, Maria Martinez, Michael Randall, and Thomas Lawn okay am i in the correct view here yes looks great okay great um um so we gratefully acknowledge the native peoples on whose ancestral lands we gather as well as the diverse and vibrant native communities who make their home here today in Maryland so in this talk i'm going to be describing some of the largest and finally worked obsidian artifacts uh described in the conference including results from portable XRF as well as microbeam XRF in the scanning electron microscope and i want to thank all my co-authors on the study especially Maria Martinez uh from the National Museum of the American Indian who introduced me to these special objects and Meredith Sharpe-Noyes, my former postdoc who took the lead on the measurements and did an excellent job during her fellowship at the Smithsonian so i want to begin by drawing a distinction between the assuredly pre-columbian round and polished objects that are found in various places in Mesoamerica and those in sharp distinction with the rectangular polished objects that are the subject of the top today that have never been excavated or otherwise found in context to our knowledge um and these objects are part of the NMAI collection here's the the building that you may be familiar with on the national mall in downtown Washington DC and they're curated at the on behalf of um indigenous communities at the cultural resources center in Suitman Maryland just outside of DC so for a variety of reasons that are spelled out in Maria Martinez's chapter in a forthcoming volume that's going to be published by the Smithsonian scholarly press these objects are thought to be produced by indigenous artisans on behalf of European invaders during the 16th century expansion so there are five black tablets the largest one measures 41 centimeters in width 32 centimeters orthogonally and about three centimeters thick that's pretty good thickness for all of them there's one mahogany tablet reddish brown brachia and they're they're very rare artifacts and there's a small number of museums and collections in the United States Corning Dunbarton Oaks University of Missouri and in in Paris at Musee de l'Homme and at the Louvre and the two specimens at the Louvre serve as canvases for paintings by Marilla and these have been analyzed by Pixie and Caligaro and others reported in 2005 that those both those tablets at the Louvre and the five unpainted tablets at Musee de l'Homme are associated with the Ucureo Zinnipaquaro source so in this segment of our study our goal is to simply identify the obsidian sources for the six Smithsonian tablets which is the largest single collection that we're aware of and it's rivaled only by Musee de l'Homme. So given the suspicion that these objects related to the Spanish in some way we focused our attention on the major sources for obsidian production in northern Mesoamerica in the late post-classic and we've modified a map here by Mike Glaskock's 2005 publication showing the boundary between the Taraskan or Parachipa Empire and that of the Aztec with the high quality sources as we've heard a lot about in the last two days from Ucureo and Pachuca. So many of the results I'll be reporting here today have recently been reported in a Journal of Archaeological Science reports. The majority of our reference specimens come from the National Museum of Natural History specifically from the Robert L. Smith Obsidian collection which was an accession not that long ago. We have a single specimen from collected by Jose Aguilera who was the co-founder and first president of the Geological Society of Mexico and even though he was a paleontologist he donated a specimen of Pachuca to the we also received generous donations of reference specimens from Mike Glaskock and Dan Hewlin. So here are the locations of our reference obsidians from two large volcanic centers in the trans-Mexican volcanic belt. In the east we have Los Azucres and to the west Sierra de Pachuca and so we have latitude longitude information for the majority of our references and a few of them are indicated by locality name. Okay so I'm going to talk about the instrumentation we used in this study relative to what I'm referring to as a typical portable XRF and I used the tracer 3 silicon drift version because that's the one we have in our laboratory. It's got about a 5 millimeter spot size. It comes into contact with the specimen and they're all rhodium sources all of these but has a has a power of about 1.2 watts and quantification through the calibration curve method that everyone seems to be using. In in our study the portable XRF is the Elio made originally designed by XG Labs and is now a broker product. It's capable of imaging 10 by 10 centimeters. It has a 1 millimeter spot. It's a non-contact measurement and it's a 4 watt source. We don't do quantification at the moment. We're not an obsidian shock. We analyze all kinds of materials so we don't have the standards to cover all kinds of compositional ranges. So we're using a crude net count ratio method for the portable XRF. The SEM based micro XRF is also a broker product called the X-Trace. It has a lateral spot size of between 30 and 35 microns and we operated at high vacuum or in low vacuum often in low vacuum because we're not interested in putting evaporative coating on electrically insulating samples like ceramics and obsidian and so forth. It's got a 30 watt power and we do quantification through fundamental parameters. So this is what the instrument looks like on the outside. It's an Etachi SEM with a large chamber. The X-ray detector is by broker and it's a large solid angle detector. Here's the micro XRF source that I referred to earlier the X-Trace. The optic is on the interior. We also collect multi spectral cathode luminescence in this instrument and we anticipate we will be making more discoveries as we link the emission of trace species in ceramics to XRF intensities. I've termed this instrument the analysis tandem the analytical dual beam microscope and that's why you see a bicycle. This is what it looks like on the inside of the chamber and here's a schematic that's co-location of the electron beam with the photon beam and then the common X-ray detector for both methods as well as an electron image showing the polycapillary optic which is the key to positioning it. You get a complete analysis of light to heavy elements and your analysis selection is informed by electron imaging or backscattered average atomic number imaging and so you can cope with heterogeneity at the sub millimeter length scale. So our lab is designed to handle all manner of museum objects from metals, minerals, paint cross-section, ceramics, historical and pre-historical manufactured objects you name it and that's the reason we're interested in this capability. So why bother measuring lozi elements in obsidian? We all use the mid to heavy atomic number elements for geochemical fingerprinting and I would say that since the big four oxides silicon silica, lumina, soda and potash make up 97, over 97 percent of obsidian it's useful to measure that but you can also predict whether a magmatic composition is alkaline or sub alkaline. This is something geologists have done for over a hundred years by computing normative mineralogy whether a magma will be quartz free or quartz bearing and that will determine its position on a total alkalized silica diagram. Here I'm not showing all the rock names at lower silica contents than in the rhyolite field but if we project that boundary you can see that for the specimens we're interested in in central Mexico just based on the major elements from the electron beam we can rule out the patruca source. The here the obsidian tablets are the filled triangles relative to a few other sources here and we'd like to compare the not just the results that we get but where is the information coming from by these two methods from the specimen. We'll start with the electron beam and so we can perform Monte Carlo simulations under the conditions of our analysis and we find that to 99.9% of the emitted x-rays there are all for the all the elements in major elements are coming from the top 2% of the specimen. With respect to XRF we can compute the information depth for a number of elements here I've done it for all the elements you see here in the periodic table and defined as 90% of fluorescence that's the cutoff I've selected so for sodium aluminum and silicon well we're within a factor of three of the electron beam so those should be almost directly comparable especially for a pre-homogenized material like obsidian we get to the potassium you know we're up the factor of seven greater in depth and then and when we get to iron at about 70 microns that's when things you really get either you can view it as an advantage to the XRF by getting a more representative bulk number or less sensitive to inhomogeneity so iron manganese around 70 and then as we get to higher energy x-rays rubidium strontium yttrium zirconium and niobium we are at many hundreds of microns depth of the information we're getting into the spectrum and so especially for something like zirconium we're at over 750 microns we're no longer in the realm of classical micro-analysis we're somewhere doing something like a sub-millionalysis so we use a yellowstone obsidian cliff's riotic glass for this project it's an excellent reference and the reason it's an excellent reference is that back in the 1970s it was analyzed by wet chemistry by one of my predecessors Eugene Jarosiewicz so he published his work in a freely available but totally obscure Smithsonian series and a few years after he published it he got permission to republish it with geostandards and geoanalytical research it's become a highly cited source for people especially in the electron probe business who depend on the Smithsonian standards for sort of ground truth with the wet chemistry so we use it as a secondary standard as to check the quantification from both methods as well as an internal reference we we just heard in the last talk about aging factors of instruments over many years we checked this reference during each analytical session for the project you know lasting over about a year or so here's the composition from that 1980 paper of Jarosiewicz and a spectrum from the micro XRF this is unfiltered so when we compare the measurements for the electron beam induced quantification which is based upon standards just as the electron micro probe community does and with the with the XRF measurements quantified by fundamental parameters here are the results the results are quite comparable for the yellowstone rhyolitic glass there are only two oxides that exceed three sigma indifference here and that's iron tough to see because the size of the symbols are large but this is largely due to heterogeneity in obsidian i'm using an example here from one of our tablets but you get the idea that there are iron oxides on a fine length scale and the micro XRF beam is a factor of 30 times smaller than the PXRF so we have the ability to encounter differing iron values by virtue of the the size of the probe so for silica these are the two values the XRF value the air bars here are one sigma and the electron beam value so they're just outside of three sigma but when we take a look at the accuracy of the analysis we're going to compare these results to the wet chemical values we we find that we're in good shape we're both the electron beam and the micro XRF results are within one and a half percent so that's absolutely we're very happy with that level of accuracy so here's an element ratio biplot for the XRF net count rates for all of the all six obsidian tablets and here we're plotting iron manganese to rubidium zirconium and the data for the tablets are plotted in the blue symbols and they're largely consistent or with the ucarea source there are two tablets that have values that overlap with the zinopequero but if you were to take the mean ratio of those two values they're they plot they plot clearly within the ucareo region they're the one specimen the mahogany tablet does not align with any of these source regions in this analysis and I just want to mention that because we didn't use a calibration curve we validated the net count ratio by using well established by measuring well established references that appear in the literature and so we compared our data to three other methods in AA XRF and LAIC PMS and we had good agreement with with all of those now we were unable to assign a source for the Smithsonian's mahogany tablet based on portable XRF but there's another mahogany tablet out there at the University of Missouri Museum of Art and Archaeology and it's been analyzed by portable XRF and it's also been linked to the ucareo source much like the obsidian canvases at the Louvre and the and also the tablets from Musee de l'Homme and so this is a Jacques Stella painting it's from 1630 and it depicts the decapitation of Saint Catherine of Alexandria and it's pretty vivid and instead of blood streaming there's a there's milk coming out of the body and it's a whole thing the artist makes really good use of the of the black parts of the obsidian I notice and and then the nanoscale reddish brown areas just serve as sort of a it's part of the gruesome angelic imagery so for the next phase of our project we did several things to address the unassigned data we added two additional Los Azufres reference specimens from the national rock and ore collection we employed the SEM based micro XRF and this time around in analyzing we used all five we use five elemental ratios the five that I show you here and on top of that used a canonical discriminant analysis and we confirmed that indeed in 2D and using two of the factors in CDA we could confirm that the black tablets are indeed from ucareo there's a few points lying outside of the 50 reliability in the two-dimensional plot but when we add a third factor then they plot clearly within the 50 reliability volume for ucareo and importantly the mahogany tablet data falls within the our locality called Haruaro which is just one of the geological sources collected by Robert L. Smith and was fortunately in our geological reference collection so finally final note here I just wanted to compare the elemental ratios that were calculated from fundamental parameters micro XRF with the portable XRF and so the solid black lines here are the one-to-one relationship you'd expect if you had perfect agreement between the two methods and you know for iron manganese and rubidium zirconium we get that perfect agreement and it's notable that all those elements are present at greater than 100 parts per million we don't get that result across the board for yttrium zirconium we do for when yttrium is present at high concentrations but when it's present at tens of ppm the trend is the ideal trend isn't found and so we sought to you know figure out which of the two methods was causing this excursion so what we did was we measured two NIST reference standard materials SRM 610 and SRM 612 that had yttrium and zirconium concentrations at roughly 450 and 440 ppm and what we found is that the micro XRF gave values that from difference from expected of less than 8% regardless of that concentration of this reference standard and we got good values for the portable XRF at the high concentration at hundreds of ppm but not for the low concentration SRM and so it's clear that the portable XRF net count method breaks down for our system the way we're using it at the tens of ppm level and it systematically underestimates the concentration so that's important for elements at least in the obsidian we were looking at for well then I'll just resummarize here the I wanted to close by saying that I think the co-located electron and micro XRF beam method it looks good for analyzing museum objects but it's going to shine for uh samples that have greater heterogeneity than an obsidian and we found that the five Smithsonian uh the black tablets were linked to Ucareo and that was true by pxrf and we confirmed it with the scm-based micro XRF and we discovered a new source for the Smithsonian mahogany tablet um just 20 kilometers northeast of Ucareo and uh in within the Los Zufres complex thank you very much thank you Edward are there uh questions we have about 10 minutes before uh our for our poster session goes up so we can spend this uh asking questions uh those of you who have a poster you're more than welcome to move in to those rooms and uh prepare your poster it's awful late in some parts of the world yes yes it is and we we uh we sympathize with that I confirm you so Nico just posted uh the link to the posters in the long pdf and a place that you can comment outside of the the formal conference here if you want to uh go ahead and make any comments and beauty posters uh using those links um otherwise you can move around the rooms and view the posters and talk with the authors directly which we certainly welcome but we we understand some of you it's very late and you may not have the energy for it we understand that one of our poster presenters will not be available but I think we're going to post the we're going to share the poster either myself or Kyle might might post it um just so people can read it and if there's any questions we can copy those questions down for the authors and then convey those um to them well I have enjoyed uh your conference so far thank you very much yeah when we next group of talks we've been very fortunate to have so many um stay on persistently throughout the shape that it's a dedicated community the high silica community