 So to begin with, I would like to thank the organizers to have accepted the following communication. The present topic is particularly meaningful for me because I have approached it since my master's degree, working on early medieval glass from the Eastern France, and I currently finished thesis on glass from the 8th to 11th century in France, in both typological and archeometrical aspects of this material. This presentation will deal with phenomena prior to the abandonment of natural ice flocks and the collapse of the centralized production model of antiquity. During the last two decades, advances in the chemical analysis of archeological glasses have established an increase in recycling in Western Europe prior to the adoption of plant or wood ash as alkalis at the end of the 8th or in the 9th century. However, archeological and archeometric data have also demonstrated that raw glass was still used during this period to produce vessels and stained glass windows. This raw glass was produced into two main regions during the antiquity as Elevant and Egypt. I will focus my communication only on 11th-century glasses in France. The material you all considered here comes from 10 sites. The glass samples are dated from mid-7th century and the mid-9th century. I would like to point out that this corpus is not representative at all because it is not the result of a systematic approach of early medieval glass in France. We cannot analyze all the glass remains found on the territory for several practical reasons that you can easily imagine. The data most results of opportunities to access to the archeological material and interests of the context. So, the number and the distribution of the sites do not reflect any reality except the state of research. Among the 10 sites, there are two domestic occupations. Navdaim in Azaz is a real world settlement mostly based on agricultural and domestic activities like weaving. Only nine fragments of glass have been found in early medieval context. The second site is characterized by the new urban occupation in the city center of Marseille close to the hairball. Among several late antique vessel glasses, a blue-green pan cup with twisted ribs belonged to the late Levantine glass group. Three sites are more characterized by the religious occupation represented by a church or monastery. Even the domestic or funeral structures have also been recognized on these sites. The glass material studied here comes from religious buildings. That is the case of the funeral church of Luxei, a Crombenius foundation and the one built at the top of a fortified settlement close to Salah-Leban in the Jura Mountains. On both of them stained glass windows have been discovered in late Maryvingian early Carolingian context. The last site also gives up several lamps and vessel glass with probable liturgical function. The third religious site is an early medieval monastery founded at the top of an hill in the Volge Mountains. It's the Salmon. For more than 1,000 glass shards are associated to the first occupation dated from late 7th to early 9th century. Stained glass windows, vessel glass and lamps show great diversity of colors, ornaments and shapes. Among them, we find characteristic glass with reticula walls. Five glass workshops complete our list. The oldest is Magalons I on the Mediterranean littoral attributed to the 6th, 7th centuries. Only one shard shows all the characteristics of the late Levantine glass. In Bordeaux, the excavations of the Camille-Julienne Square revealed several raw glass chunks and shaves of crecibles, but in secondary position. One of the chunks comes from the Levantine coast. Another raw glass chunk and workshop wastes have been fortuitously discovered around the famous Abbey of Jumiège. The archaeological material is tupu to identify the production but two stained glass windows and a small piece of vessel glass completes the batch. The glass workshop of Amage shows the same configuration. Different modules of crecibles and glass modules have been discovered in a large trash level between the church and the living building of the nuns. In this case, the production of reticula glass has been demonstrated by both analysis and typological data. Evidence of glass workshop has also been found in Salon-les-Bains and Luxeuilles that I already presented, but there is no proof that Levantine glass has been locally remelted, contrary to these workshops of Jumiège and Amage. The last site represents a different type of workshop, not directly connected to religious environments. In Meru, in the north part of France, the craft structures are implanted along the river and closed to essential resources like clay, wood and sand. Two main glass groups have been identified. The first one is a soda mineral glass made according to the antique recipe and the second one is a mixed alkali glass resulting of a mix between natural glass and wood ash alas. Levantine glass was identified among the culettes, those recycled glass. The ceramic material allows us to date the workshop into the first two thirds of the 9th century. The glass samples have been analyzed by ICPMS. This method allows us to determine almost all the major mineral and trace elements of the glass. With the laser ablation, we can also control and choose to analyze a specific part of the sample like the ornament or the body of the cell glass. Let me now present you the results. I will detail the first steps of the classification which consists to separate the glass samples according to their flukes and then to identify relative flue glass among natural glass to the recycled ones according to copper, antimony, lead and stain oxides amounts. This graph shows you that we can easily make the difference between the two main origins of raw glass in late antiquity and early Middle Ages. The titanium and zirconium amounts of the Levantine glass are low compared to the Egyptian one. In the current state of my research, I did not identify late Egyptian glass in France among the stained glass windows and glassware. Following the classification proposed by Phelps and Ali in 2016, the glass identified as uncontaminated by recycling turns out to come from the eastern Mediterranean coast. The late Levantine glass is mainly and easily identifiable by the lake of Manganese contrary to the raw glass produced during antiquity in the same region. That's why its presence above 500 ppm could significate that it's not natural and that a few proportions of culette could have been remelted. So we propose to apply to more strictly selecting excluding the samples with more than 500 ppm of Manganese. Focusing on this new selection, we can compare the French samples to the two known late primary glass workshops of Betel Yeze and Apollonia. Both of them are located on the Levantine coast in the actual state of Israel. It's interesting to notify that most of the samples dated between the mid 7th and the mid 9th century belong to the Apollonia group while this primary glass workshop is attributed to the 6th, 7th centuries. It concerns the raw glass chunks of Gignage and Bordeaux, the glass waste of Gignage and Amage and most of the vessel glass and stained glass windows are found on the other sites. Only one sample, a stained glass window from Amage in the north of France is similar to the Betel Yeze primary production. Two other ones from the same site could belong to the same group that's there on the intersection of the two groups so it's nepotism. So how to interpret these results? At the end of the Merevingian period and the beginning of the Carolingian one before the transition toward raw-glass glass chemical analysis have demonstrated that pristine Levantine glass is still worked and remelted in French secondary glass workshops to produce stained glass windows and vessel glass. But the main observation is that most of these Levantine glass seems to come from the glass workshop of Apollonia. A difference of one or two centuries appears between the primary production and the secondary one. So we propose two explanations. The first, there is actually a long interval between the primary production in the Levant and the use of the glass chunks in France which could be explained by the recycling of older stocks of raw glass. The most of the Levantine glass discovered in France shows similarities with the Apollonia group but it could come from primary glass workshops still unidentified by archaeology in the same area of Nonswan and which use the same primary resources like sand. Only an increase of chemical analysis and new archaeological discoveries could confirm one of these proposals. In any case, we can make a last observation on the use of this pristine glass. It doesn't seem to be reserved to a specific production. Furthermore, the glass boroughs did not seem to give importance to this resource because they did not hesitate to mix it with recycled glass. So Levantine glass was probably not a precious resource during the 7th, 9th centuries even though it was less common that recycled glass was collected. Thank you for your attention.