In all the years of our trying to build tree-ring chronologies for the Eastern Mediterranean, two of our biggest headaches have been the Roman and Hellenistic periods. Hundreds of buildings survive, many with beam-holes in them all of which are (usually) empty. Stroll around the Baths of Caracalla in Rome, for example, and you will see hundreds of places where beams once upon a time were. Now, at best, they are full of pigeon nests. Until this year, most of our Roman wood had come from pilings chainsawed from the River Kupa in Pannonia at Siscia (modern Sisak or Celtic Segestica).
In the summer of 1996 our Roman wood supply increased dramatically when we acquired 42 timbers from near the port at Herculaneum (first photo) with the kind help of Dr. Mario Pagano, the Deputy Soprintendente, and Jo Berry of the British School in Rome. In the vaulted Roman cellars was a carpenter's workshop stacked with timbers, some showing signs of use and re-use. When nearby Mount Vesuvius erupted in A.D. 79, they started to burn, but the burning was almost immediately extinguished by some 30 meters of ash and tephra which fell on the site, protecting it until the present day. The reducing atmosphere (lack of oxygen) was the equivalent of a charcoal kiln, and the beams, almost all of them pine, with occasional cypress and a bit of oak, were splendidly preserved. We now have a 202-year chronology with an end-date no later than A.D. 79.
In the current Italian excavations (Scavi Novi) adjacent to the Villa dei Papiri
a deep trench has revealed Roman buildings, preserved up to the roof, also buried
under some 30 meters of lapilli, ash, and tephra. From one of these we collected
another dozen large timbers which crossdate with the material from the
carpenter's shop. The last existing ring in this trench is 18 years earlier than
the last one in the workshop. Since the bark is missing in both cases, we do not
have a difference in felling years between the one lot and the other, but this
is a promising start, and we trust that future collection from the Scavi Novi and
other locales will reinforce and extend our new Roman chronology. (In photo
Scavi Novi foreman and front-end loader bring us
another burned Roman timber, freshly unearthed.)
At nearby Pompeii we were able to collect burned wood from the House of the Chaste Lovers (Casa dei Casti Amanti). This was not as long-lived as the Herculaneum wood, but it, too, is pre-A.D.79. We will have to wait for more wood to appear at Pompeii. What is most exciting about the work at Herculaneum and Pompeii is that, although the destruction date is well known, it is sometimes difficult to determine when any given building was built. We hope to be able to provide answers to these and other questions as collection and measurement continue at these two sites. It was particularly gratifying to see that every scrap of evidence at Herculaneum was being saved for future reference. Now that our Italian colleagues know what we need, there is every hope that we will be able to add to the 1996 work.
Over thirty-five years ago Michael Katzev excavated the 'oldest Greek ship,' a
Hellenistic merchant vessel which sank off the north coast of Cyprus near Kyrenia
sometime between 310 and 300 B.C. (This image is used with permission from and thanks to the National Geographic Society.) Dr. Katzev estimates that the ship was about
70-80 years old when she went down. Some planks are from the original ship;
others are from patches, or from patches to the patches. Last summer, after three
years of accumulating the appropriate permissions, we were able to bring back
pieces of 32 planks from the Kyrenia wreck. They were all soaked in PEG
(polyethylene glycol), so we had to dissolve at least the surface layer so that
we can see the rings.
Work on the Kyrenia wood is going on as I write. Two of the planks measured this week had 131 and 124 rings, and I thought I spotted another at the time of collection with 150 rings, so there is hope that a useful chronology will emerge.
We now have (from Herculaneum and Kyrenia) over 330 years of information from the
1st-4th centuries in two long sequences where before this summer we had only bits
and pieces. You may wonder why I sound so excited about this, but until now we
have had 100 times as much Bronze Age wood as we had Roman or Hellenistic. This
is REAL PROGRESS!
Few sites are able to produce dendrochronological material from periods as widely
separated in time as Medieval and Early Bronze. Sos Tepe, a remarkably
well-stratified mound about 30 kms. east of Erzurum, currently being investigated
by Dr. Antonio Sagona and his team from the University of Melbourne, is just such
a gold-mine. We have finished assembling long chronologies from both the Iron Age
and Hellenistic periods, and the assumption is that sooner or later we are going
to be able to put the whole chronological sequence from Iron Age to Hellenistic
to Roman to Late Antique together. The significance of this kind of collection,
aside from providing absolute dates for the periods in question, is that we will
then have the link to earlier periods that will give us absolute dates on purely
dendrochronological grounds.
Patrons of the Project should have received this last autumn offprints of an
article published in Nature and co-authored by Bernd Kromer of the University of
Heidelberg, Sturt Manning of the University of Reading, Maryanne Newton,
Christine Latini, Mary Jaye Bruce, and me, on a chronology which we think extends
from 2220 B.C. to 718 B.C. (Our Patrons tend to move all over the place, and mail
can often go astray. If you did not get your copy, let us know, and we will do
something about it.)
Many Patrons sent in newspaper clippings with some of the stories generated by
the Nature article, including the Los Angeles Times (dutifully repeated in both
its Japanese and Korean versions) which got it backward, saying that we had
micro-rings in 1628 B.C. and following rather than some of the biggest rings we
have ever measured. [Porsuk, by the way, now has 54 trees showing this growth
spike rather than the 36 reported in Nature.] The best post-publication story
was a testy letter to the editor of the Daily Telegraph in England (spotted by
the ever-alert Sturt Manning) from the editor of the Guinness Book of Records,
saying that he had already published 1628 as the date of the eruption of
Santorini/Thera and why were we wasting our time? Moral of story: never get
interviewed.
LHIIIB pottery is also to be found on the Kas/Uluburun shipwreck excavated by
Cemal Pulak of Texas A&M University. The last ring at Uluburun is 1315 B.C. [Note: please see the
1997 report for an emendation to this date.] I do not think we are missing many rings on the exterior. Since the shipwreck is a
time-capsule, a date for the wood dunnage on board helps date all the rather more
glamorous cargo items from half a dozen civilizations and cultures.
You can't win 'em all (or at least I can't). I reported cheerfully in two
previous newsletters that I thought we had the Tatarli Tumulus near Dinar (near
Afyon) nailed down to the sixth century B.C., although the overlaps with other
chronologies were short. I did hedge a bit in 1995 by saying that I preferred to
leave the date tentative until we have more Iron Age wood to corroborate the
so-called 'fit'. Well, my optimism has turned out to be unjustified, and caution
was the correct note.
Dr. Bernd Kromer in Heidelberg has just completed a long exercise in radiocarbon
wiggle-matching. The wood at Tatarli (with the bark present) was cut in 451 ±5
years B.C. It was then painted with what look like Archaic figures on Attic
pottery of the 6th century (see newsletter of December 1991 for the sketches but
ignore the date given there). Note how flat the radiocarbon curve is for these
centuries (7th-4th B.C.). Without wiggle-matching, 14C dating is a fruitless
exercise for the Late Geometric through Classical periods.
A question which nobody as far as I know has been able to document (other than
guess) is: how long does it take a style/fashion to percolate from an urban
center (say, Athens or New York City) to the remote provinces (say, Afyon or
Ithaca, N.Y.)? Here at Afyon/Dinar/Tatarli we seem to have a painting style that
was popular in Athens a half or three quarters of a century earlier.
We spent a week on Santorini/Thera in 1996 exploring old mining quarries for
evidence of organic material at the 'contact surface' between the volcanic tephra
and the Minoan soil. Joan Ramage (Cornell geology) had spent a week with
Professor Floyd McCoy (Hawaii) on a surface survey of the island, and Dr. Sturt
Manning (Reading) joined us for several very hot days of reconnaissance. We found
imprints of now-deteriorated vegetal material, but nothing substantial. We did
find nuggets or nodules of what appears to be highly-concentrated sulphur.
Chemical analysis of these might help show us exactly how sulphurous the Theran
explosion actually was.
In the laboratory, Mecki Pohl has now finished Version 20 of his CORINA (Cornell
Ring Analysis) Program as part of his "Diplom" thesis at the University of
Berlin, and Jennifer Fine and Laura Steele have finished converting an enormous
amount of tree-ring data to it. The working copy of their latest Bronze Age/Iron
Age Chronology is 1598 years long, composed of 446 data sets with 55,132 rings,
each one of which was measured at least twice. The number of intervals (between
years) with four or more samples (even as many as 108) per year is 1375. Of
these, 591 intervals or 43% have trends where 75% or more of the ring-widths for
a given year increase or decrease together (= "signature" years = German
Weiserjähre). For instance, 1081 B.C. was not a good year for Eastern
Mediterranean trees: 1 tree has a larger ring than in 1082 B.C.; 66 trees have
smaller rings. Presumably the wheat farmers had much to complain about in 1081.
For readers who remember the Rhys Carpenter thesis about a climatic contribution
to the end of the Late Bronze Age, 21 of the 35 years following 1159 B.C. (the
presumed date of the eruption of Hekla III) are starred "signature" years, with
ratios of increase/decrease in any year ranging from 30/0 to 0/37. (English
translation: in one year 30 trees put on larger rings and 0 did not; in the other
year no trees put on larger rings and 37 put on smaller rings.) Something really
interesting climatologically is going on when 100% of a tree-population scattered
over many kilometers of terrain is responding in an identical manner. More on
this later when we finish figuring out what the patterns mean.
Jennifer and Laura have also continued to supervise the activities of fifteen
people in the lab. Laura's thesis on Neolithic architecture (much of it written
last year while she was in the hospital) earned her a summa cum laude. She is now
working for us full-time. Mary Jaye Bruce continues to produce quantities of
readable text and figures as well as keep orderly accounts in six currencies in
five languages. She also was a stalwart of last summer's collection trip from
which we returned with almost a quarter of a ton of samples from some thirty
sites in five countries. Maryanne Newton and Christopher Roosevelt (who is almost
finished with his M.A. thesis on the Amuq pottery excavated by Prof. Robert
Braidwood about sixty years ago) won PhD Fellowships from the Graduate School at
Cornell, so we expect to see them around for some while to come. Miles McCredie
is our systems analyst/programmer/trouble-shooter when Mecki is out of the
country. See our World Wide Web site at: http://www.arts.cornell.edu/dendro/ for
more information.
Summer menu improbabilities: CHICKEN SPIT and HAT WOTHER
Most unlikely form of transportation: ALBATROS AIR
And don't miss: VERTIGO BEACH with its CLUB POLLEN.......
Peter Ian Kuniholm
SOS TEPE: MEDIEVAL AND HELLENISTIC AND IRON AGE AND EARLY BRONZE AGE:
THE LONG BRONZE-AGE/IRON-AGE CHRONOLOGY:
MASAT HÖYÜK, KAS/ULUBURUN:
Additions to this long chronology include the upper Hittite level at Masat, the
wood from which was collected years ago from Prof. Tahsin Özgüç of Ankara
University. It dates to 1353 B.C. (no bark present, however, so the cutting date
could be slightly later). This is of interest not only to scholars of Hittite
history but also to scholars of Mycenaean history and archaeology because the
building in question has imported Late Helladic IIIA/B stirrup jars, a well-known
class of pottery. In the photo is a Hittite room at Masat with
Hittite and Mycenaean (LHIIIB) pottery. Dendro date is 1353 B.C. (no
bark). Note imprint of roof post on rear wall.
KUSAKLI HÖYÜK
Another dated building is the Hittite temple at Kusakli excavated by Dr. Andreas
Müller-Karpe of the University of Regensburg. Photographed from the air, the temple has a last preserved ring (no bark,
though) of 1384 B.C. (photo credit, Prof. Müller-Karpe). A cuneiform text fragment is thought on palaeographic grounds
to be from the 13th century B.C. which would mean that the building was either
rather long-lived or a number of rings have burned off the exterior of the wood.
We expect to be getting more examples from Kusakli later this year. It will be
interesting to see whether they bring the construction date down significantly
from 1384. [See: A. Müller-Karpe, Antike Welt 1996 fascicle 4, 305-312 for further
details and more stunning photographs.]
IRON AGE:
To our long Bronze Age/Iron Age chronology we can now add the pine sequence from
Ayanis, the Urartian citadel and palace on the east shore of Lake Van, built by
King Rusa II. The last ring (with the bark) was formed in 651 B.C. Since Rusa
reigned from ca. 685-645, this means the place was built shortly before his
death. The building inscription on a single block of basalt, published by Altan
Çilingiroglu and Mirjo Salvini, listing Rusa's accomplishments, leaves one-and-a-half lines in mid-text blank. Apparently, Rusa was planning to do other things
at Ayanis but never got around to them. [See: Çilingiroglu and Salvini,
"'Rusahinili in Front of Mount Eiduru': The Urartian Fortress of Ayanis (7th
Century B.C.)" Studi Micenei ed Egeo-Anatolici 35 (1995) 111-124.]
IRON-AGE TATARLI (A CAUTIONARY TALE):
THE NEOLITHIC:
Maryanne Newton finished her M.A. thesis on the dendrochronology of Neolithic
Çatal Höyük. The sequence as we currently have it wiggle-matched (ten accelerator
determinations done at the University of Arizona) runs from ca. 7024 B.C. to ca.
6449 B.C., not much different from the provisional results (based on only two
determinations) reported last year. The bargraph below shows how the sequencing
works from Çatal Höyük Level X to Level II End. A summary report on the Çatal
results will appear in Dr. Ian Hodder's On the Surface, published by the McDonald
Institute and the British Institute of Archaeology at Ankara, currently in press
at Cambridge and due out any day now.
MORE ON THE 'CAN'T WIN 'EM ALL' THEME:
MEANWHILE, BACK AT THE RANCH...
END NOTES...
CORNELL UNIVERSITY
Back to Index