Posted by: viscenter | August 26, 2009

EDUCE Project Highlighted

3DscrollsThe Vis Center team working on digitally unrolling Herculaneum scrolls through the EDUCE project was recently highlighted in an article in the Lexington Herald-Leader. Check out the story.



  1. I’ve been following this project for several years; this is very exciting, if painstaking, work.
    At this stage I was wondering; do the “assembly programs already exist so that you can begin making trial test runs? Or do you have to write the basic software to recreate two-dimensional pages and then see if there is enough contrast, vis a vis the ink vs papyrus, to read anything?

    Even if you have programs, perhaps derived from medical settings, I bet you would still have to massage them quite a bit.

    And even “cataloguing” the data in an efficient, machine readable form sounds very daunting.
    Good luck and Best Wishes

  2. We have semi-automated software that works quite well on test objects. We can “unwrap” simple folds and view those surfaces as flat images. What is harder is the unpredictable curves and breaks of something like these scrolls. So in that case we’re working with tools we’ve built that let a user identify by hand the exact position of where the correct layers are. This is more tedious, but gets around solving the (hard) problem of finding all the layers correctly and automatically.

    • Your efforts are the greatest gift to classical scholarship in perhaps a thousand years. I’ve been watching your progress only since June and while my Latin is OK, I am learning ancient Greek in anticipation of your successes. Is there a daily or weekly blog so as to keep up with your work? I check the news almost everyday for updates but there is very little out there. How can classical enthusiast help? (write our congress people and try to get more EDUCE funding?)
      Wish I were there!

  3. Still very early in the process (and semester). I do impatiently wonder: Have you identified whether the ink is carbon-based yet on this scroll. Thus, will contrast problems drive you crazy, or will it be reasonably possible to eventually puzzle out the document in a reasonable time period?

    I don’t suppose that the little identifying tag that extends out of some of the scrolls survived on this one. That would truly be too easy.

  4. While my enthusiasm, which is nearly boundless, does not transit the boundaries of personal competence enough to learn classical Greek as does Mr. Rudder’s, I am incredibly fascinated by the explorations you are conducting.

    Indeed, Tom Palaima’s website at the University of Texas does offer incentives to fantasize about learning to read Linear B, and therefore Mycenean (and classical) Greek, enough to be tempted. But the idea of truly earning a language with a synthetic rather than analytical syntax seems a summit too high for this observer.

    I know one fears premature announcements, and taking positions one may later have second thoughts about, but you did have the courage to post the initial images of the interior structure of the scrolls.

    Perhaps a little sense of the hard work and frustrations you may face in interpreting the data may lend a little more excitemenbt to this website.

    By now you must have some hint, at least, if the contrast issue (composition of the ink) is going to be a major problem. is it a carbon-carbon situation, or not. If not, I’m sure, as you indicated, that there may be many creative if demanding solutions.

    Surely, by now, you may have a few guesses. we realize that we are in media res right now.

    Good luck.

  5. If ink is unreadable;
    A spectral analysis of the carbonized ink compared to a spectral analysis of carbonized papyrus might yield a differentiating chemical signature that the scanners could focus on. An interesting intellectual exercise if nothing else.

    • I see that this has already been asnwered:
      International Journal of PIXE (IJPIXE)
      Volume: 18, Issues: 3-4 (2008) pp. 279-284
      “… that elemental distributions from the ink used in this scroll mainly contained Al, Fe and Ti as well as minor trace amounts of Cr, Cu and Zn”


  6. Each letter of text then is a field approximately 15,000 microns by 10,000 by 1 micron, characterised by Al, Fe, and Ti. With the SkyScan 1173 resolving down to 15 microns it’ll be tough to get that third dimension. Is there enough density in a 1-micron thick carbonized sample to be picked up in the other two dimensions?

  7. A thought just popped into my head.

    I don’t know if anyone checks this site other than Mr. Rudder or myself, but why not throw out another question that surely the experts know the answer to.

    Perhaps influenced by Hollywood movies, I have always assumed that only one side of a scroll was written on. You know the image of the town crier standing in forum, scroll in front of him, reading the Senate’s latest edict, or some such.

    But it just occurred to me that this might be a little inefficient. If one were writing books, extended essays, etc., one could work oneself down (across?) one side of the scroll, flip it over, and then read the other side. The only reason not to do so that leaps out to me would be that the curl of the paper, especially as it ages, might make this difficult.

    So the question. Did the copiers of these volumes use one or two sides of the sheets of papyrus? If both sides, I guess that would only make the process of transcription that much more demanding. Certainly the amount of information would be doubled per scroll, but in addition, in some cases two faces of characters would be touching one another. This last would also relate back to Mr Rudder’s one micron query.

    I can only assume one micron thickness is enough for discrimination because the journal article (to which I have no access) had to have been written before this experiment was begun.

    But two characters touching one another? I can conceive of success, but indeed a challenge. Of Course readers might have made notes about the text, important in themselves, on the obverse side, or, of course, in the margins.

    On a totally different track. It just ocurred to me to wonder if somewhere in Herculaneum there were municipal records archives, tax collections, receipts, contracts, land use, water and sewer line records, was there a land title registration system? Maps/surveys? Census data. City council minutes. And if so, was there any possibility of those records being likewise preserved. Some of this stuff had to exist for a town to operate. Pure luck if any of it survived, I presume.

    I suspect that question will be answered, if ever, in some excavation long after I have shuffled off this mortal coil.

  8. Typically the writing appears on only one side of the scrolls. This is known from the hundreds that were taken apart (with varying degrees of success).

    As for the existence of Herculaneum records, etc. – the manuscripts that were discovered in the Villa could be pretty much anything. The chaos of the Vesuvius eruption led to evacuation, but not everything made it out. One really intriguing idea is that there may be another wing of the library still buried containing more manuscripts.

    Yes, we read and maintain this site and appreciate engagement with those who are interested in our project.

  9. Thank you for your response. Glad to learn that my unexamined presumption was correct.

    Yes, I had read about the speculations concerning another “library,” perhaps a “Latin” library in another room (The lost books of Livy!, my fantasy would be Claudius’s History of the Etruscans, which from Suetonius’s comments surely included a chapter (“book”?) on the history, structure, and word stock of the Etruscan language. He describes the young Claudius journeying to interview old women (native speakers) in mountaintop villages gathering information and data. Maybe he would have earned an Anthro PH.D. in a modern context? rather than be forced so rudely into politics late in life.)

    Fantasies aside, I have read that there are the thousand or so unread scrolls, which I gather everyone assumes comprise second-rate Hellenistic musings on philosophy, theories of art, etc. (no histories, lost poems from the Troy cycle? Sophocles? Never mind).

    This surely would keep scholars busy for mucho time if you are even marginally successful in your endeavors.

    About which, any hints?

  10. The ability to discern individual letters or portion there-of and their relative position to adjoining letters is one of the first things I would have sought from the data. I would suppose, risky as are all suppositions, that by now the answer to this question is known and not open to spurious speculation. Whether, that answer is yes or no, the techniques being developed to combine all the data into the best big picture will make future efforts either easier as a framework for even larger volumes of data from finer scans or employed assembly line to open the ancient library to everyone.

    Sometimes the stylus used in ancient times left indent marks in the papyrus, the papyrus being perhaps 200 microns thick, the thickness or depth of any such mark having too various determinants to presuppose. Any evidence of this would help determine pigment application methods employed or may offer suggestions of lettering where ink is gone.

    Having very little experience building in virtual realities, I suggested to a friend of mine, Lillian, who is a skilled builder in “Second Life” that she could build this scroll there from screen-capture slices of the fly-through video. Once built, it could be enlarge and studied, walking around inside as-it-where. She said sure, then I told her about Dr. Seales and this effort, and she said Uof K had the best resources for this and just be patient. Seeing as how my friend’s efforts could not include lettering anyway…

    Which led to the thought of a final 3-D model, with color enhanced lettering? Perhaps, a fly through model of only the Fe atoms? Daydreamed of flying through concentric cylinders of bronze greek letters floating in a black void. Seems like a cylinder model might be an intermediary step toward the actual unrolling.

    As we consider the possibilities of the scrolls contents I would add the off-chance that we may have 1800 copies of the same treatise. Writers of a certain rank would have kept many copies to give sell, or trade. Their location, when found, only suggests they were important to some one. Why so we may yet come to learn. I am more optimistic than this, but tempered and again there is the suggested adjoining library room(s).

    I do recognize the many restaints to the projects ability to quell, answer, or entice these speculations.

    Looking forward to enlightenment!

    Mr. Rudder

  11. Thank you again Mr. Rudder. Your comments are very helpful and enlightening.

    As to scroll contents, like you, I doubt that an 1800 scroll sales inventory would still be around a century after the writing date, but who knows. I find the physical size of the library astounding in what would appear to be a private establishment.

    I don’t truly have much hope that my fondest fantasies, the lost books of Aristotle (as opposed to his lecture notes, which we have), lost plays of Sophocles, etc., or my special fantasy of an Etruscan grammar will emerge. But 1800 scrolls by “mundane” writers on “mundane” subjects would still be literally a treasure beyond compare.

    [As to what can be teased out of very rudimentary texts, check out Tom Palaima and the PASP website at the University of Texas. He specializes in Linear B, most of what survives are inventory and tax lists Knossos, Thebes, and especially Pylos. They were inscribed in 13th century (BC) Greek in a syllabic alphabet on thin clay tablets; Palaima thinks as notes for later entry on some sort of paper or parchment general ledger. None of them appear more than a year old at the destruction event, most a few weeks or monthsold. The fire that destroyed the palace baked most of the tablets in (at Pylos) what appears to be a transcription room. Still an enormous amount can be learned about the economy, political structure, names (Achilles the blacksmith, Nestor the goat herd, that sort of thing) deities worshipped, provinces administered, even the names of the King and military head at the moment of destruction. Think what can be learned from volumes of fully formed sentences, thoughts, quotes, evaluations, even offhand references. For instance, cities on the island of Euboea (sp?) were giving tribute to Thebes, which has implications for political power, though no mention of King Odysseus]

    At this stage, I would consider “success” on this project to be the ability to read a few lines of text to identify one or both scrolls, and perhaps the beginning of the arduous process of assembling, the text. Not necessarily its completion. That will surely involve the development and refinement of software, hardware, and skills/literally “the art” in the old usages of craft and engineering. From this they will surely also learning how to enhance the collection of the imaging data in the future.

    So, however irrationally impatient, I’m very hopeful for the future; whatever problems they run into now.

    Sorry that I seem to run off at the mouth.

  12. Oh, your comment that Greek verbs are hard? I’m not surprised, but I’m also curious about the nouns and modifiers. The language seems so highly inflected.

    It’s clear from scanning some of Palaima’s articles (which also make passing comments on linguistic change over the six or more centuries preceding classical Greek (which of course his audience implicitly understand) that there appear to be a hell of a lot of case endings (all with names mysterious to me), far more than the four in German. I can’t conceive what the syntax must be like. I guess this may have been the norm for Indo-european languages about 3,000 years age, though. All enough to daunt me, as beautiful as it is to listen to a few lines of Greek poetry or to a Greek play, I know I could never at my age attain fluency to understand/to feel the original.

    But I certainly applaud your efforts.

  13. We still have not completely unwrapped/flattened any of the data to be able to fully examine what we can see in the scans. But we have assembled definitive data sets for each scroll. These sets are made up of multiple segments (blocks) that overlap. We had to carefully be sure there were no gaps or overlaps (redundant slices). We have 4 sets for each scroll at different power settings and resolutions. Looks like our highest resolution scan is about 15 microns; the complete scans are captured at about 27 microns.

    There are intensity variations in the data that are intriguing and there are the following possibilities that we are exploring (but haven’t answered definitively yet):
    1. Can see all the ink (this would be fantastic, and due to impurities may still be a possibility)
    2. Cannot see regular ink but can see special inks/pigments that may be present and are different in composition from the regular writing
    3. Cannot see ink but can only see structure

    [the “structure” is plainly visible – density variations in the structure of the papyrus]

    Now that the data sets are together and confirmed correct we are removing the data from the container that held the scrolls during scanning and are preparing to “unwrap” the most promising (and easiest-to-segment) layers.

    We are moving slowly because we want our work to be correct, definitive, and systematic.


  14. Fabulous.

    Thank you for sharing this

  15. This paper, on multi-spectral imaging of ostraca (pottery shards with writing and notes on them) may perhaps be of interest to persons browsing this site. Multispectral imaging studies of the Dead Sea scrolls are also discussed.

    Of course the physics of taking a photographic/digital image using a wide variety of frequencies of reflected light is quite different than what is being attempted here (as I understand it, exciting molecules to give off em radiation within an object which then can be detected and digitally manipulated into images) the article does illustrate the exciting and multipronged exploration of subjects for imaging studies as well as the variety of techniques being developed and explored.

    One supposes in particularly difficult cases in the future, one might apply MRI-like techniques to supplement the other imaging techniques, perhaps for objects one might never before have considered as bearing epigraphic content.

    Here’s the citation.
    Bearman, G. & W.A. Christens-Barry. 2009. Spectral Imaging of Ostraca. – Palarch’s
    Journal of Archaeology of Egypt/Egyptology 6(7) (2009), 1-20. ISSN 1567-214X. 20
    pages + 22 fi gures.
    Keywords: ostraca, spectral imaging, archeological imaging

    Found it on today’s Archaeology News.

  16. I think, Dr. Seales’ statement is worthy of a press release.
    As great a boon as having clearly defined data is having data that clearly defines the question (or possibilities).

    The statement regarding the data on the inks and the possibility that other chemical properties (impurities) might prove indicative, took me back to the PIXIE analysis, where it was initially surprising not to see oxides in the chemical lists of the pigment. From this I had assumed the carbonization process occurred in an oxygen depleted environment (the rooms sealed off from air by ash and magma). Other avenues of contamination within the scrolls would seem to be few…humidity perhaps?

    The density variations are encouraging to see (if I could see them ). My relatively simplistic understanding is that each of the 8,000 slice-scans would present themselves as a two-dimensional data set with “pixels” of 27 microns indicative of the density at any point. By tying together, around a defined vertical axis the 8,000 horizontal slices, patterns of density would occur in the vertical dimension giving definition to pages, pigments, etc. Are specific densities representative of specific chemicals? Having reviewed the fly-thru-vid again there are portions of several ‘pages’ that stand-alone and are obvious targets for ease of mapping. The outer parts of the scroll are more tightly wound and suggest shrinkage due to the heat; in these parts the distance between “pages” might again be only a handful of microns and mapping more difficult. I’ve tried and been unsuccessful to determine in what language the SkyScan 1173 spits out data, or even operates in that I might better grasp the task at hand.

    I’ve also tried, and thus far been unsuccessful in, determining the depth of a reed-pens scratch in papyrus from evidenced examples. My statement from last semester reeks of slip-shod academics, I invite citations of scholarly results on the issue.

    One of the earliest known fragments of the Iliad is from a pottery shard. From these earliest fragments some of the evolution of the Homeric legends can be deduced. Seeing the CT scan of the scrolls holding container made me wonder what could be gained by CT scans of pottery; first thought is authentication, second is text written in under coats of the paint, thirdly, potters technique and materials, and finally ,of course, two thousand year old finger prints.

    My final thought for the new semester is the possibility of other libraries, commercial-scroll-sellers, or copy-shops in Herculaneum and to lesser degree in Pompeii. Only a fraction of Herculaneum has been excavated. A commercial scroll seller might have a greater diversity of authors and subjects and being more populists by nature perhaps most interesting (I know I’m spuriously speculating).

    Most of my statements here are regurgitant of Dr Seales work, or speculations built upon his work. The best perspective I can find is that of teacher-student. In that perspective I apologize for my stupidities but not my ignorances, and do seek correction on both counts from all.

    Mr. Rudder

  17. I also found the following last fall:
    “Greek Papyri – An Introduction”
    Edited by E.G. Turner, Clarendon Press
    Good broad discussion of the various papyri issues. As an intro text it might be redundant to these readers.

  18. It is interesting to speculate just what books or documents would be likely taken away in an precipitous evacuation such as at Herculaneum. My guess is not very many, if any. Personal mementos or perhaps an original manuscript in the hand of a famous individual. I doubt most people would have the presence of mind to find papyri in the midst of impending disaster. If the victims had coins or jewelry with them that may well have disappeared during the Bourbon excavations. Apparently capsi were found collected in a manner to suggest they were going to be moved for safekeeping. Pliny the Younger’s description of the eruption suggests that conditions deteriorated so rapidly that few escaped. Pliny the Elder died apparently from fumes and ash and he had only tried to approach from the sea and had not reached the boathouse area, where all the bodies were found in recent times.

    Could it be that this was a room dedicated to the memory of Philodemus, a sort of museum? The question is whether 2,000 odd scrolls too many to likely be attributed to only one writer? That does seem like a prodigious output. Perhaps we could get lucky and some of these papyri are of great works that Philodemus liked and collected.

  19. Herculaneum Roll 1570
    Yesterday I found Joseph Ponczoch’s Master thesis (BYU) describing roll 1570. While this one was partially unrolled circa 1800 into seven major fragments Mr. Ponzoch description of the structure, lettering, et al provides a greater insight to what we can expect to find in these EDUCE project scans. Mr. Ponczoch also demonstrates the type of analysis all these scrolls will go through once the ink is discerned before they can be called readable.
    Here’s a link to his paper:
    Mr. Rudder

  20. Hopefully the unopened papyri will have conserved much more of the text.

    I guess there is no chance of any Latin papyri.

    • Dr. Knut Kleve, in his report to the 20th International Congress of Papyrologist, describes 62 or so Latin papyri found at Herculaneum as “heavily damaged, first by mud and water, which washed out most of the text, and next by carbonization. The difficult state of the Latin papyri is why they have remained almost unknown to this day.”
      He goes further to describe Pherc 817 as being a fairly well preserved example.
      Dr Kleve then describes 30 as written in Early Roman Script, 11 in Pre-Classical Capital Script, and 17 in Classical Capital Script.
      Complete two page article: 20th International congress of papyrologists: Selected papers By Adam Bülow-Jacobsen, pps 382-383 (also available online)

      There may yet be some real Latin Gems to be unrolled here!

      Mr. Rudder

  21. Mud and water damage – that’s interesting. Perhaps there had been some pre-tremor problems in the Villa with the water table or maybe these were papyri that were on their way to being discarded, maybe too worn, and dumped in a leaky “basement”.

    Was there enough time for the ink to be washed out before the pyroclastic flow? If they were covered with water they would have been boiled and then carbonized. Could be ct scanning would document water damage to the fiber structure of the papyri.

    • I’m thinking;

      Any glaciers or snowcaps would melt with the eruption and those hot waters would be the leading edge if not the body of the pyroclastic flow. The flow is estimated to have come down at 300 degrees Celsius and flows can reach a top speed of about 450 mph (variable depending flow density, etc.).

      The Philodemus Project out of UCLAs classics department states that some papyri were found in a capsa near the entrance to the long peristyle and others on wooden shelves in a room just inside the entrance to the villa from the long peristyle.

      The Latin scrolls may have been a bit dated at the time of the eruption (variation in script suggest a span of a century or two) and it’s easy to suppose that they might have been on the bottom of the shelving where flowing water would have damaged them more so than those on upper shelving, and then again when/if the shelves collapsed those scrolls in the lower levels would be most heavily damaged by it’s debris.

      Those in the capsa outside might also have been damaged by flowing hot waters as they would have had no building structures to protect them and once water filled might have soaked the papyri and washed it’s ink away, boiled as you say. I am more inclined to think the better-preserved Latin fragments came from the capsa due to its inherent better protection, but again I’m speculating. Freshly written new Greek is as likely to have been in the capsa; trying to save Phildemus’ latest best-seller?

      Documentation of the early digs is available in Italian via the 8 volume “LE ANTICHITÀ DI ERCOLANO ESPOSTE” which is posted in it’s entirety online (but no translation). I don’t read (or speak) Italian but details of the papyrus findings (such as they are) might be found there. Any Italian readers volunteer to translate?

      Mr. Rudder

  22. The Romans must have been aware of the shorcomings of a permeable ink. I am sure wine, etc. must have gotten on papyri from time to time. I wonder if the ct scans will reveal the composition of this ca. 79 ink. Iron gall was developed later, I believe.

    • The PIXIE analysis showed the carbonised ink consists of Al, Fe and Ti as well as minor trace amounts of Cr, Cu and Zn. I too have assumed a plant based ink as iron based comes much later.

  23. As we speculate here I’d like to add a list of identified authors from The Bodleian Collection of Herculaenum scrolls and scroll fragments:

    Caecilius Statius – 1 Piece
    Carneiscus – 1 Piece
    Chrysippus – 1 Piece
    Colotes – 1 Piece
    Demetrius Laco – 9 Pieces
    Epicurus – 11 Pieces
    Lucretius – 1 Piece
    Metrodorus – 1 Piece
    Philodemus – 49 Pieces (Others credit 36 scolls)
    Polystratus – 1 Piece
    Scriptor Epicureus incertus – 1 Piece
    unattributed – 42 Pieces

    My piece count is not meant to be definitive just a study in expectations for the scroll Dr. Seales and the PROJECT are working on.

  24. That composition would at least theoretically bode well for “reading” the lost text.

  25. Pigment part of the ink was Carbon – Lamp-soot
    This chemistry would probably be the binder which made up about 40% of the ink. Variety of ink binders available at the time, tree saps, plants, one boiled crushed fish to a syrup.

    Beyond curiosity of technique;

    Metals show up in X-rays, right?

  26. I am guessing on this but I would think you could look for known metals in the ink using the appropriate spectrum

  27. Dr Seales came to the same conclusion several years ago. The PIXIE anlaysis cited in my Nov 1st posting is a spectral analysis of the ink. Since the pigment is carbon these chemicals must be the binding agent.

    Has anyone ever done a chemical survey of ancient Greek/Roman ink-binding-agents?

    Might be useful to have such an inventory as specific virtual-unwrapping-techniques might work better on different chemistries than others.

  28. In spite of all the historical explanations for their disappearance I still find it hard to accept that virtually no papyri(or even codice) have survived from the Roman era. Or that no one tried to “bury” manuscripts for posterity during the Decline and Fall. Of course the truly hidden would only be found by accident, but I have never hear of anything like that, not even in Egypt, where the chance of preservation are the best. Dumps don’t count – I mean a trove that was intentionally stashed away from barbarians and religious fanatics. Discouraging.

    Some scholars have suggested that some major works, like the Satyrica of Petronius, could have actually survived to as late as ca. 1100. I am not so sure; I think the monks may have copied everything that was available to them – a great part of the literature may have been already lost by 500 or so.

  29. Where the Papyri were found and some scroll counts?

    From: Ethel Ross Barkers’ “Buried Herculaneum”, Adam and Black, 1908

    Barkers’ sources are mostly the letters of Camillo Paderni, keeper of the museum of Portici (which received many of the early treasures), and the reports of Karl Jacob Weber, the Swiss engineer on the site from 1750 thru 1764. Weber documented and mapped many of the finds in the initial Villa tunneling. Weber’s efforts are well documented in Christopher Parslow’s book “Rediscovering Antiquity”.

    The following excerpts from Barkers’ book are worth quoting:

    In October 1752 the first papyri were found in the tablinum. 21 volumes and fragments contained in two wooden cases. [Tommy note; cases here means capsa].

    In the spring of 1753, 11 papyri were found in a room just south of the tablinum, and in the summer of the same year, 250 were found in a room to the north.

    In the spring and summer of the following year, 337 Greek papyri and 18 Latin papyri were found in the library. Nothing of any importance was discovered after this date. The numbers given here exclude mere fragments. Including every tiny fragment found, the catalogues give 1756 manuscripts discovered up to 1855, while subsequent discoveries bring the total up to 1806. Of these 341 were found almost entire, 500 were merely charred fragments, and the remaining 965 were in every intermediate state of disintegration.

    The German, Sickler, who worked on the rolls in the Bodleian from 1817 to 1819 under a Commission of distinguished men presided over by Lord Castlreagh, destroyed seven of the rolls [Tommy note; while trying to open them].

    The original library-Though it is difficult to reckon accurately, the original Library probably contained something like 800 volumes, as far as we can judge from what remains, though the actual number of works was much less, since a volume of papyrus never contained more than one “book” of a work. Many of the works were in several volumes; for example, Epicurus On Nature numbered no less than 37 books, and the library contained three copies of it. There re also two or more copies of several works, so that the actual number of works in the library may be reckoned at something less than 200.

    The Latin manuscripts.-Besides these Greek manuscripts, 21 Latin papyri were found, nearly all together in a wooden case (capsa), as well as many fragments. They are so ruined by the damp as to be almost undecipherable. [Tommy note; I think these are the October 1752 finds]

    The locations of these finds are mapped at a google-sponsored site “AD79 Destruction and Rediscovery”
    The library is described as having wooden shelving along the walls with a double row of outward facing cases down the center and a small bust of Epicurus (12 cm in height). The shelves are described as having cornices of design similar to those contemporary to the 1750s dig. To my knowledge the library is no longer accessible due to backfill of the tunnels. Some of the backfill may contain papyrus fragments as workman early on trashed the scrolls not knowing what they were.
    Mr. Rudder

  30. Virtual “Villa of the Papyri”

    Researching where the papyri were found has led to recognition of the need for virtual reconstruction of the Villa. Not only as it was in its heyday, but as it is now buried in rock. Probably many archaeological sites will ultimately be documented this way. Any grad students need a project?

  31. Well, my last two posts have been shot down by the moderator,
    this one will be gone soon enough as well,
    for that brief moment:

    Thank you to everyone in the Project for the path(s) of inquiry your example has sent me down.

    Cheers & gby,

    • I apologize that I missed approving your previous comments. I missed the e-mail notification saying you had posted them. I have approved them now.
      Thanks for your interaction on the blog.

  32. Research Commentary

    While looking for a copy of Michele Ruggiero’s works (he published many of Weber’s drawings), I stumbled upon a five page dissertation from 1879, in Italian, a scanned pdf image from an original published source.

    I found this citation via Google books, was there able to download the Italian into a Microsoft Word-Doc which I was then able to translate with Google translator, ultimately to find the dissertation wasn’t relevant. All in less than hour.

    I compare this to my undergraduate efforts of some years ago when a citation would be followed up through inter-library loan and two months later hard-copy in hand I’d have gone begging to an Italian student to give me a sense of it. Two to three month effort to learn its irrelevance.

    Technology is sweetening research and education!

  33. Reviewing the history of the digs I find some relevant points;

    1.) Paderni, in his letter to Hollis of 1755, mentions having spent 12 days removing scrolls from the library and that finally some were too fragile to be moved and were left. And again according to Winckelmann, in 1754 Paderni had decided it was scientifically impossible to open the scrolls so that he considered it better to leave them in the excavations.

    2.) Piaggio, in a later letter, says that many scrolls were damaged when workman directed seasonal water flows through the tunnels to help washout the dug tuff

    3.) Dr. Sickler in a letter dated 1817 reports several columns of text were destroyed in his trunk while traveling from Naples to Paris. The one surviving half column he did translate and publish. Subsequently he destroyed more scrolls trying to open them.

    4.) A large room or area in front of the door to the library has apparently never been fully excavated, nor the personal quarters. Weber and Paderni had lost interest in the villa by about 1761 when the scrolls had proven too difficult to open and more statuary wasn’t forthcoming (and it was easier to dig Pompeii, it being closer to the surface).

    5.) A small pile of flour was found by workers in another building on a table and reported to Paderni, whose trained eye spied ink and letter structure among the dust and identified it as the remnants of an uncarbonized papyrus.

    6.) James Porter in his “Hearing Voices” dissertation at the University of Michigan mentions that 80 scrolls were broken when they were moved to Palermo in 1798 and returned to Portici in 1802.

    There is then a very good chance for substantive future papyrus finds in the Villa; in the library and scattered on the floor just outside the library’s door. Paderni’s characterization of those papyri he saw and left suggests that techniques for scanning the papyri “in situ” might need to be considered.

  34. I could certainly conceive of finds which could not be disturbed and whose information could only be preserved by an “in situ” scanner. I am optimistic about portable equipment, but will there enough data or information left as to be decipherable even by room-sized machines?

  35. Intensity variations in the data sets, is Dr. Seales phraseology. That statement gives some pause for optimism. I’ll be happy to see those variations mapped.

    To which I’d add; 15,000 Germans last year studied ancient Greek, that’s 15,000 computers in Germany that could be put to this task in a SETI-type analysis of the data blocks. Double, triple, Quadruple that number for US, UK, Italy, and Greece. We’d have every intensity variation in every scroll mapped within a few years.

    • Wouldn’t it be wonderful if their were some degree of the intensity variation that was indicative of a chemical that was unique to one of those found in the binding agent of the ink?

  36. “Intensity variations” could mean that information is preserved in the scrolls even if current technology cannot ferret it out. Who’s to rule out that future forensics won’t be up to it.

    Meantime we should be good stewards and try to locate and protect whatever bits and pieces of the papyri that might have been left about in the 18th century. The Bourbons would be absolutely incredulous at the idea of ct scans. It would appear to be like magic to them. So let’s not rule out future wonders of recovery.

    • Conservation of the papyrus is difficult. The 200-year-old hand drawn copies in the Bodleian tomes are today showing detail that no longer exists in the originals (where originals still even exist at all). Just as the scans being done today will show lost detail in years and decades to come. One Univeristy attempted to conserve papyrus many years ago by sliding them into plastic sleeves-which has raised an entire new set of problems some 40 years later.

      One of the biggest points of conservation is getting those out of the ground that are there now. Beyond those already speculated upon, the latest excavations suggest up to four levels at the Villa leading down to the old seashore. The hope is that they will find many scrolls, boxed up (I mean capsa’d up) on one or more of those levels, ready for evacuation. Philodemus’ was well-read and if this Villa’s papyrus is his collection the literary jewels of classic Greece and early Rome may be sitting 30 feet or so directly underneath current excavation efforts.

      Success in the virtual unrolling of these burned scrolls will add a great deal of muscle to encourage the Italian politicos to open the site to more digging and thus help preserve those not yet found.

  37. One more item regarding the chemistry of the scrolls:

    In an article entitled “Bookselling before the invention of the press.” published in “Living Age” magazine, Volume 20 page 313, June 1845, previously published in Chambers Journal;

    “Each scroll was usually washed in cedar-oil, or strewn between each wrap with cedar or citron-chips, to prevent it from rotting or being eat’en by insects”

    Some of the densities seen in the scans might then be charred cedar chips. I don’t yet find any other reference for Citron chips being used this way.

  38. Biosciences company Organovo has developed a 3-D laser printer for bio-technologies, first goal being the building of blood vessel. The device builds 3-D structures with a fineness of 20 microns.

    This device, in theory, could build a 3-D model of the scroll such that it could be physically touched and manually unrolled.

    Can the SkyScan 1173 data be fed into the Organovo product or something similar?

  39. A detailed 3-D model of a scroll would certainly be impressive but I don’t know if it would be necessary to extract text. You wouild need an adequate three dimensional map to select the right data for that section of the papyrus. If the segments were small enough curvature and folds wouldn’t seem to be such a problem. The 3-d map would take care of the flattening in the reassemble of the segments

    What really intrigues me is possible scanning using heretofore unavailable frequencies of radiation, but which can now be generated. In particular it would be nice to have beams that could penetrate kinda like a neutrino. I am thinking of that “lead burrito” coffin that was recently unearthed not far from Rome. This a variation of the “in situ” challenge.

  40. The April 26th video is such a beautiful slip of papyrus. Classic image of what one would expect.

    No lettering is jumping off the page screaming at us, but my off-handed guestimate was that this technique would yield 2-4% of the original text image (non-carbon binder being 40-60% of the ink, thickness of ink lettering being 1 micron, scanning fineness being 27 microns). When this image is examined with the expectation of seeing 2-4% of the ink there may or may not be evidence of lettering, hard to tell. Larger segments may reveal more useful contexts.

    This technique definitely warrants further development with finer scanners as they become available.

  41. I’m currently reading some the Philodemus work by Dr. Obbink and he refers to chunks of scorze containing several layers of charred papyri that haven’t been opened.

    Perhaps for technique development, focus and testing on the smaller scorze might be more fruitful than a full scroll. I read Dr. Seales comment that it took half an hour each time just to recalc, these smaller targets should mean a smaller data-mass. A large number of such fragments hold important clarifications for lines of text.

    Revelations of the internal structure of the papyrus might also aid placement of the fragments which is often problematic for these smaller pieces.

  42. There is one other quality that separates the carbonized ink from the papyrus and that it its reflectivity when presented with visible light at certain angles. This suggests there is a fundamental difference in the texture of the surface of the carbonized ink versus the papyrus.

    I’m not sure how this quality could be exploited by the methods being employed but it does seem to lend some hope of a possibility.

    I do still favor development of a scanner with 1-micron discernment, as that should lift the text from the page. For faster computer processing I can only think of NASA and their cloud-computing model. Although again I could see the volume of data from one scroll even overwhelming that system.

    There is answer to this…

  43. Quote from Space article Feb 2, 2011

    “Researchers at the Center for Free-Electron Laser Science in Hamburg, Germany, say a new X-ray laser may let them watch individual molecules in action during processes from brain-cell activity to photosynthesis.”

    “The new Linac Coherent Light Source X-ray laser, which came online in 2009 at the National Accelerator Laboratory in Menlo Park, Calif., is so much brighter and faster than previous lasers that researchers hope the technique may reveal molecules interacting in their native habitat.”

    This might be the beginning of the answer…

  44. Is Terahertz scanning the answer? I was reading the attached story and noticed they had a similar problem with detecting carbon ink overlaid by another layer of carbon that X-rays didn’t pick up but spectroscopic terahertz imaging did.

    Couple useful quotes from the article:

    “…Terahertz waves can be reflected from various layers within a painting giving a three-dimensional picture of how the artwork was put together. At the same time, terahertz waves are selectively absorbed and reflected by different types of pigments and materials and this gives spectroscopic information about the nature of the paints that were used.”

    “…And since the atomic weight of carbon in the signature and in the surrounding canvas and paint is very similar, an x-ray would not have picked up the difference. ” ~implies their technique did?

    “…The spectroscopic capabilities of terahertz imaging ought to allow accurate identification of pigments and other materials used in the painting.”

    “…it’s not hard to imagine that art historians around the world will want to get their hands on terahertz imaging machines in the hope of identifying other unexpected features in famous paintings.”

    food for thought….

  45. Does the SkyScan 2140 Micro-CT / Micro XRF Scanner, offer enough improvement?

  46. Reading an article on funerary ancient masks being scanned in South Dakota for hidden text, can’t help but wonder if the techniques developed here might be applicable:

  47. U.S. Department of Energy (DOE)’s Lawrence Berkeley National Laboratory (Berkeley Lab), a collaboration of researchers used low energy or “soft” X-rays to image structures only five nanometers in size. This resolution, obtained at Berkeley Lab’s Advanced Light Source (ALS), a DOE Office of Science User Facility, is the highest ever achieved with X-ray microscopy.


  49. X-Ray phase-contrast imaging.

  50. Congratulations to Dr. Seales and the UK team as the fruits of yesterday find application in the solution of today!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


%d bloggers like this: