 Thank you. Well, thank you for accepting our paper. And the mouthful of this title, just the important words here are disassociated and multivibrate classification methods. And also, as a general disclaimer, when this paper was submitted, it was only four of us. I am presenter O, author, I'm Jocelyn and I am a chemist. And some of my research is on dice. Marian Weldon is the head of the Green Mark Special Collections. I'll go back to that. Carl Buch is also a chemist, but he deals with multivariate classification methods. And together, both Carl and myself, advise Amelia Miller, who, and this is what happened when we submitted the paper, but things changed. And Amelia got married and now she's a million speed. And two other wonderful women joined the project, Caroline Chen and Olivia Yeager, who are on their last year of college and they are on their way to become also part of heritage scientists. And to entice you, with the presentation, I'm showing you just five of the textiles that we had the opportunity to analyze. And they're not to scale. All we know about them is that they're from the Andes. That's it. And the three institutions that are those who are involved in the project are the Winter Tour Museum, who is not this winter tour in Switzerland, but winter tour in the United States. The University of Delaware, where pretty much everyone put Marianne is appointed to, and Bryn Mark College. Everything is in that star within two hour radius. In case someone here doesn't know where we are, you drive two hours north and you hit Canada. And if you keep driving south, you hit Mexico, which is where I am from. And I just like saying that all the time. For the interest of the project, it's relevant that we know what kind of collection the Bryn Mark College has. Bryn Mark College is a small liberal arts college. And because of the very small size of the institution, all of the students have potential access to everything in the collection. So you can see how small the college is. It's less than 2,000 students. So everyone can potentially have access to the textiles. As a result, this collection that we were dealing with is classified as a study collection. And the only thing that we know about the wonderful textiles we were dealing with is that they were donations and they came from the Andes. Hence, there are no links to the excavations. We don't know the exact location. We don't even know the approximate location. And the only thing that we know is the one that someone who came in the 2000s and someone that name we don't have and classify them by eye. If you like the textiles that you are about to see, please take a picture or copy that link because in there you will find all of the textiles. But going back to the collection, this is in the background very faded on my left is Marianne Weldon who is using part of the collection. This picture is actually from a Coptic textiles analysis class that she taught. And in an email she wrote, we have never been able to link to Donor's travels in their archives publications or in speaking with their adult children and grandchildren to their collecting these textiles. We believe they purchased them from a dealer. Again, the dealer we don't know the name, as most are cutting the rectangles or pretty small bits. Also, most are sent to paper when they came to Green Mark College. I realize that this is a horrifying picture, cutting the textiles into bits, but this was a very common practice in the early 20th century in the United States. Wealthy people will go abroad, purchase a bunch of textiles. Let's say purchase a bunch of textiles. And if they didn't fit the mantelpiece the way they wanted, they would cut them and frame them. So it's not uncommon for this to happen. And that's pretty much what we know about the textiles. And as a result, these disassociated textiles have no research value. And it's really sad because no one was using these wonderful textiles until we came. The things that pretty much everyone in the audience I'm sure knows is that we face a lot of challenges when dealing with textiles. They are contaminated and not just from the original burial context, bodily fluids and things in dirt, but especially textiles like these after excavations who knows what happened to them. Maybe they were washed, maybe they were fumigated and along, etc. Of course, we expect them to be degraded. So maybe we are not even aiming to identify everything that they may have. If we can have samples as everyone here already said, the sample availability is going to be very, very, very small. And the last but not least thing, again, everyone knows, we are likely going to deal with extinct materials. So what happens if we have a very large group of seemingly degraded, seemingly contaminated textile samples that have limited sampling availability and that probably have the same extinct materials. If we throw the beauty of statistics into them, we may be able to identify trends, which is what we were doing. In case someone here doesn't know what this means, the multivariate analysis means I will use you as an audience as an example. If I'm doing statistics and I choose to do multivariate classification, I can choose very specific variables. Let's say we add the variable of chosen profession. That may not leave a lot of trends. But what happens if we add archaeologists specifically or textile conservators or chemists or along, etc. That would give me very specific variables that I would be able to fine-tune the statistical analysis that I want to do. And that's pretty much what we did. But before telling you about the results, I think it's fair that you bear with me for two slides because the next two slides are related to the inspiration of this actual project. So for the past five years of my life, I have been dealing with very new textiles. These are late 18th, early 19th century British textiles specifically from Norwich in the East Anglia region of the United Kingdom. And amongst the aims that that project has is to develop a non-destructive methodology using one of the techniques that Joan mentioned. And what we did, we went into these sample books that are in pristine condition. We did liquid chromatography and we also collected reflectance spectroscopy. And then we threw statistics at them checking if it was possible to just use reflectance spectroscopy and non-destructive analysis to identify the chemical markup. And these are just a few of the samples that we have been able to collect. And in this particular case, we have only yellows and we can see the statistical analysis that we have three groups. So thinking this is non-destructive analysis, we have access to a very large collection of disassociated textiles. We inquired at the Greenmark College about their study collection if we could do something similar. And that's what we did. So in the picture, you can see Olivia Yeager, Caroline Chen and then Amelia Miller that it's there. So you can also ask her things. And we just went to Greenmark College, took a bunch of textiles and collected reflectance spectroscopy or force. As John said, it's great because it's non-destructive. It's highly portable. To give you an idea, you can put it in your carry-on and actually check it at the airport. It's really fast in less than five minutes. You can see if the analysis is working or not. It's highly reproducible and it gives us a quite broad range of spectral information that ranges from the ultraviolet to the near-infrared. So we can also have vibrational information. And it's very sensitive. It's sensitive enough to possibly see things like some of the chemical microbes that we were aiming to see. We collected the samples and then we did statistical analysis. And we did a specific style that is unbiased statistical analysis. Now, when I explained very briefly what the multivariate classification methods was, I chose very specific variables. So I was biasing our data in a very specific way. But if we throw all of the information and tell the software to do statistical analysis by itself with the help of a human behind it, it will find any trends if they exist. It will find trends only if they exist. So that's exactly what we did. And we chose to just start with yellows and reds because among other things, they have been analyzed. Keep in mind that I'm showing you just two colors, yellows and reds. Yet, we find very easily three groups here. So the questions are, are they related to hues or are they related to chemical markup? And because we were really hoping it was related to the chemical markup, we went back to Greenmark College and this time we did micro-destructive sampling, liquid chromatography. But this is a very large number of textiles. In our case, we chose only 18 textiles and for each, we took two samples, one yellow and one red. It's very small in case someone here doesn't know. The sample that we took, it's about half a centimeter. So for someone my size, 159 centimeters, 5'2", for those who live in Imperial, it's half of my pinky nail size. So it's quite small and one of the things that we do in the group is do twice the extraction to be sure we didn't miss anything. That's one of the practices that we have. So we compare the data with the statistical analysis and one of the groups was just reds. It was to probably no one surprise, Cochinio. And then the other two groups were yellows. One, we chose to call grasses. It had a pigenin and some of the chemicals that are present also in European grasses. And others were curcetin-based. The two plants that I'm listing there are referenced from the wonderful paper of Ilaria de Gano and co-workers at the Colombini Group. And we checked that one of the groups in the yellows was related to just grasses and the other one had pretty much all of the curcetin-containing plants. So a different plant material group. And then we did something even crazier. Keep in mind, this is just statistical analysis. We were not touching again the textiles. So we decided to throw some statistics with early 19th century, late 18th century textiles. Keep in mind that the chemical markups are the chemical markups. So in this particular case, we decided to do this crazy, seemingly crazy thing, because we already have a very sound group that one of the groups had curcetin, one of the chemical markups in some of the yellows. And we did what in the group is called the happiness. In green, you find the 18th century textiles. And in red, you find the Andean textiles, the archaeological textiles. And you can see there is a very clear overlap between the oval here, one of the greens, and the rest of the yellows. So that particular oval that I was showing you is the one that had little to no curcetin. And that coincided with our same results. Still, you can see the two different groups, which means we have a range of curcetin containing plants. We wanted to see if these chemical results were related to specific types of textiles. And again, keep in mind that this is not accurate, probably because it was by visual assessment only. What we saw is that we didn't find any specific trends. That lead us to our results. So we already have promising data that with this technology and with this multigraphic classification overlapping spectral information with chromatographic information, we can find trends that we can then reclass, we can use to study through a statistical analysis only other textiles. These textiles are disassociated. But the moment we find one textile that has context, the rest, the 13 to 21 textiles that we have will find a classification to a culture, to a time period, etc. We plan to continue doing this research. And even if we don't find the specific plants that they were died with, we know which non-identified plant is. So we have plant groups in specific textiles. That's promising. We will continue working on this. So especially if someone has an open excavation of India and textiles, it would be very nice to talk to you because we might be able to do something together. And of course, we're going to expand on other colors. And thank you to the museum, the University of Delaware Greenmark College, and of course the people behind it. Olivia Yeager, Amelia Speed, and Caroline Chen. So if for everything, copy my email, I will be there. And thank you.