Gunnar Heinsohn: Londinium’s Dendrochronology

Dendrochronologist Petra Ossowski Larsson has repeatedly emphasized that so far it has not been possible to link a post-Roman tree ring sequence directly to timber or roof beams of Roman Imperial Antiquity (1-230s AD):

“Primeval oaks, i.e. those that could grow undisturbed for a long time and therefore have many annual rings […] are abruptly exhausted in England around AD 200 (conventionally), then young oaks and/or branches are used for another 50 years and then one changes to stone and brick as building material”

(Ur-Eichen, also solche die lange Zeit ungestört wachsen konnten und deshalb viele Jahrringe aufweisen […] sind in England um AD 200 (konventionell) abrupt erschöpft, dann verwendet man noch ca. 50 Jahre lang junge Eichen und/oder Äste und dann geht man zu Stein und Ziegel als Baumaterial über;
email, 13 June 2018).

If we examine this finding in Londinium — the largest city in Roman Britain — it is undoubtedly also confirmed there.

Some authors may write early 3rd century AD instead of AD 200.

Yet, after — say — AD 230, no more new Roman period buildings were erected in Londinium.

Some authors believe that very few buildings from the time up to AD 230 were still inhabited in the 4th century.

But nobody claims that, up to the 10th century, new Londinium houses were built upon the ruins of Londinium houses destroyed around AD 230.

Thus, no strong trees were felled for timber and roof beams for buildings upon such ruins on which dendrochronological measurements could be made.

For timber classified as Anglo-Saxon (in the London area outside Lundenwic) from structures that do not stand on Roman ruins, felling dates are given from the 7th century onwards (Tyers/Hillam/Groves 1994, 14).

This brings the earliest Saxons to the Late Latène period, which lies 600-700 stratigraphically, but is set in the textbooks 700 years earlier in the 1st century BC (see in detail Heinsohn 2018).

The sudden end of building in Roman style and technology does not come as a surprise because a catastrophe had destroyed Roman Britain and Londinium.

Large parts

“were already covered by a horizon of dark silts (often described as ‘dark earth’). […]

Land was converted to arable and pastoral use or abandoned entirely.

The dark earth may have started forming in the 3rd century”

(Schofield 1999).


“many [British] building sequences appear to terminate in the 2nd and 3rd centuries. […]

The latest Roman levels are sealed by deposits of dark-coloured loam, commonly called the ‘dark earth’ (formerly ‘black earth’).

In the London area the ‘dark earth’ generally appears as a dark grey, rather silty loam with various inclusions, especially building material.

The deposit is usually without stratification and homogeneous in appearance, it can be one meter or more in thickness. […]

The evidence suggests that truncation of late Roman stratification is linked to the process of ‘dark earth’ formation”

(Yule 1990, 620).

At the same time, the economic heart of London stopped beating:

“The whole port begins to collapse […] by the third century”

(Milne 2016).

Because there are no traces of freshly felled oaks between the 3rd and the early 10th century for structures built upon ruins of the Roman period, scholars use the dates found in our textbooks to date Londinium for the period from the 3rd century onwards:

“Knowledge of the approximate date is used, not as prior evidence, but to save time during the crossmatching process.

All the Roman timbers, for example, can be grouped together, and time will not be wasted in comparing Roman sequences against medieval chronologies”

(English Heritage 2004, 25).

With this non-scientific method a “floating yet fixed“ (Malaga Bay 2017) Romano-British chronology had been created.

Thus, it had been decided not to test the possibility of a direct connection of Roman tree ring sequences (ending around AD 230) to sequences of the High Middle Ages (with primitive beginnings around AD 930).

But it is at this point in time that construction begins anew in Londinium — with simple huts sunk into dark earth.

It is consensus that there was

“only exiguous activity amid Roman ruins before 950” AD

(Blair 2018, 344).

London’s new beginnings mirror Charlamagne’s Aachen.

In the “second third of the 10th century”1 (Erkens 2013, 580), modest pit houses were built there into a “mouldy or alluvial layer, which everywhere overlays the purely Roman layers in great thickness”2 (Sage 1982, 93).

1 zweiten Drittel des 10. Jahrhunderts

2 Moder- oder Schwemmschicht, die […] überall die rein römischen Schichten in großer Mächtigkeit überlagert

Aachen’s dark earth proves once again that an enormous cataclysm must have wiped out Roman civilization just before the onset of the High Middle Ages.

Whatever method is used to date Londinium timber before the primitive buildings from AD 930, the researchers (astronomers, dendrochronologists, C14-daters etc.) have to justify why the Roman-era buildings were not destroyed around AD 930 instead of the early third century.

Dendrochronologically they have no timber, i.e., no new construction upon the ruins of Londinium before c. AD 930.

If they want to put hundreds of years between the assumed end a few decades after AD 200 and AD 930, they should have tangible Londinium evidence to prove this.

Those who insist on about 700 (mainstream) or about 300 fallow years (various dissidents) must be able to show windblown layers.

In such a long time, massive forests have grown whose dead roots, along with the shells of molluscs etc., must be visible.

300 to 700 years worth of plant growth will have left sediments.

You won’t find them upon the dark earth in Londinium below the new construction from 930s onwards.

Thus, the experts have a difficult decision to make.

Do they want to work with tangible evidence or do they insist on serving chronological dogma?

This question is not only addressed to experts for Londinium, but to all scholars, who want to compensate missing human habitats for complete epochs of the first millennium by ice layers, tree rings or even by celestial constellations.

Dendrochronologists can work most successfully in places that, for the period between c. AD 1 and AD 930, have archaeological strata only for the Early Middle Ages (c. 700 to 930).

One may call them site type A.

However, they are well helped if they can resort to woods (best suited are lumber and roof beams) from the 10th/11th century and later, where the historical dating is reasonably certain even without scientific procedures.

In the London area, Lundenwic (with residential structures for the 8th-10th century) belongs to this group.

So-called Anglo-Saxon timber in the London area outside Lundenwic can also be interpreted with greater reliability by connecting it to 10th/11th century woods double-checked against historical chronology (Tyers/Hillam/Groves 1994, 14).

One wonders if time was “wasted“ to compare these woods to the chronologically swimming Roman period items from Londinium (English Heritage 2004, 25).

Dendrochronologists run into problems, though, if their site of type A has a phase of Late Latène (1st century BC).

If one does not explain to dendrochronologists that this period, stratigraphically, immediately precedes the Early Middle Ages — well shown in Bachórz/Poland (Parczewski 2006, 125) — then they are in danger of searching for a date at least 700 years older than 700 AD instead of something between 600 and 700 AD.

The problems even get bigger in sites of type C where the pre-930 stratigraphic “rings” sit in Imperial Antiquity, i.e. in 1-230s AD.

That is the above case of Londinium.

Yet, the most famous example is provided by Rome.

The researchers do not “know” from stratigraphic evidence but from their learning programme since childhood, that the wood below the 930 AD layer must be at least 700 years older.

They don’t consider that the wood may only be one year older.

Instead of asking why their site has no layers for 700 years between 930 and 230 AD, they start with the firm belief that one will eventually find them.

They see no reason to think any further.

Errors are then inevitable.

Tragic is the case of the famous dendro pioneer Ernst Hollstein (1918-1988).

The Larssons masterfully proved to him that he used a sequence of 207 years twice to get closer to the supposedly much earlier Roman past.


“all dendrochronological datings done on West Roman time wood is wrong by some unknown number of years”

(Larsson/Ossowski Larsson 2010).

One must therefore explain to the young dendrochronologists that, stratigraphically, Roman civilization comes to an end directly before the Tenth Century Collapse around AD 930 (Heinsohn 2017).

Such a basic training course would focus on the comparison between buildings and small finds or coins found at sites of type A in the 9th century and at sites of type C in the “2nd” century with strikingly similar architecture and small finds.

The problem for dendrochronology is similar with sites of type B with Constantinople as the most famous example.

Again, the researchers do not “know” from hard evidence in the ground, but from their learning programme since childhood, that the Late Antiquity wood found below the 930 AD destruction layer must be at least some 300 years older.

They don’t consider that the wood may only be one year older.

Again, a basic training course would focus on the comparison between buildings and small finds or coins found at sites of type A in the 9th century and, with everything strikingly similar, at sites of type B in the “5th” century.

If prospective dendrochronologists begin their studies with comparative building research of the “2nd” [C], “5th” [B] and 9th century [A] and see the identity of techniques and materials, they will continue to ask for the stratigraphy of lumber and roof beams they are demanded to date.

Then they will quickly realize why dendrochronology is relatively easy for the 9th century, but so difficult for the “5th” or “2nd”.

That is why the Londinium specialists, supposedly dealing with timber of a floating 2nd century, have a particularly complicated task whilst the Lundenwic researchers have it easier because they can put their hands on 9th century woods that may be linkable to the historical chronology beginning in the 10th century.

If the two groups meet one day and wonder why Londinium has its dark earth layers in the 3rd century whereas Lundenwic’s belong to the 10th century, they might try to find out together if they are both researching in the same stratigraphic levels of the 8th to 10th century.


-Blair, J. (2018), Building Anglo-Saxon England, Princeton: Princeton: University Press

-English Heritage (2004), Dendrochronology Guidelines on producing and interpreting dendrochronological dates;

-Erkens, F.-R. (2013), “Aachener Geschichte zwischen Karolingern und Staufern“, in Kraus, T.R., ed., Aachen. Von den Anfängen bis zur Gegenwart. Bd. 2: Karolinger – Ottonen – Salier 765-1137, Aachen: Mayersche, 471-583

-Heiberg, J.L. (1898; 1903), Claudii Ptolemaei opera quae exstant omnia, vols. 1.1; 1.2, Leipzig: Teubner

-Heinsohn, G. (2017), “Tenth Century Collapse“, in [Quantavolution Magazine],

-Heinsohn, G. (2018), “Finding Bede’s Missing Metropolis“,;

-Kunitzsch, P. (1974), Der Almagest: die Syntaxis mathematica des Claudius Ptolemäus, Wiesbaden: Harrassowitz

-Larsson, L-A., Ossowski Larsson, P. (2010), “The ambiguous match in the Hollstein chronology”,

-Malaga Bay (2017), Deranged Dating. The Roman Problem¸

-Milne, G. (2016), The Growth of London as a Port from Roman to Medieval Times, Gresham lecture;

-Parczewski, M. (2006), “Die Forschungen an der Siedlung mehrerer Kulturen in Bachórz, Kr. Rzeszów, FSt. 16. Die Grabungssaisons 15.-19.”, l’Institut d’Archeologie de l’Université de Cracovie, ed., Recherches Archeologiques de 1999–2003, Craców 2006;

-Sage, W. (1982), “Die Ausgrabungen am ‚Hof‘ 1965“, in Cüppers, H. et al., eds., Aquae Granni: Beiträge zur Archäologie von Aachen, Köln & Bonn: Rheinland-Verlag & Rudolf Habelt Verlag, 91-100

-Schofield, J. (1999), “Saxon London in a tale of two cities”, British Archaeology, No. 44 [May],

-Tyers, I., Hillam, J., Groves, C. (1994), “Trees and woodland in the Saxon period: the dendrochronological evidence“, in J. Rackham, ed., Environment and Economy in Anglo-Saxon England, Council of British Archaeology, Research Report 89;

-Yule, B. (1990), “The ‘dark earth’ and Late Roman London”, in Antiquity: A Review of World Archaeology, Bd. 64, Nr. 244, September, 620-628;

Editorial Assistance

Thanks for editorial assistance go to Clark Whelton (New York), and Tim Cullen (Malaga).

Gallery | This entry was posted in Dendrochronology, Heinsohn Horizon, History. Bookmark the permalink.

7 Responses to Gunnar Heinsohn: Londinium’s Dendrochronology

  1. Clark Whelton says:

    Well done, Tim!


  2. This site says the baobabs that have died recently, in one case exceeds 2400 years. That should settle the matter for dendrochronology.

  3. Unfortunately the baobabs are not datable by rings, they are dating them using radio carbon.

  4. johnm33 says:

    This makes a lot of sense, and connects some dots for me. Years ago [97?] I was chatting with a couple of guys who did London walking tours who had a deep knowledge of their subject and they spoke of the continuity of the cities site use going back to ‘celtic’ times and how the city fathers had always come to terms with invaders including the Normans.
    IIRC the vintners building n/w of southwark bridge was where the temple of ISIS once stood and the continuing practice of ‘swan upping’ and the vintners guild[?] annual feast of swan following the journey up the thames to another site of a temple to ISIS near oxford being one of the stickier tales.
    This does beg the question, did the city fathers actually invite first the romans then the saxons over to assist them in the battle for supremacy between the various ‘celtic’ kingdoms? Caeser and Claudius both had native allies. It would make some sense of Londons seperate legal status too..

  5. Gunnar Heinsohn says:

    One reader asked if the Larssons hadn’t changed their mind after 2010. They had written, that “all dendrochronological datings done on West Roman time wood is wrong by some unknown number of years”.

    Yet in 2016 they wrote:
    “When the supra-long continuous tree-ring chronologies produced for climatological and geological purposes became available in the first decade of the 21th century, we were able to simply date the floating Roman chronologies. We did indeed not find them at the date predicted by the historians, but more than 200 years later. This made the Roman gap collapse.“

    For the oak chronologies with which the Larssons confirmed their statement in 2010, however, nothing has changed at all. As they openly concede, the gap is still there:
    “Within the oak chronologies [gap collapse] has not yet been confirmed because the necessary raw measurement series are still kept secret or are undiscovered.”
    (Cf. Ossowski-Larsson, P., Larsson, L.A. (2016), “Redating West-Roman history -about specious twin events and anachronisms in Late Antiquity“;

    It can’t be surprising. After all, the Larssons do not show cities with stratigraphies in which timber from early medieval residential quarters lie above layers with timber from Late Antiquity, below which timber from residential quarters of Imperial Antiquity is found. The climatological and geological experts have not shown such urban stratigraphies either.

    Gunnar Heinsohn

  6. Gunnar Heinsohn says:


    Attentive readers have asked me if I could summarize my findings on Londinium and Lundenwic once again in a nutshell.

    I put forward 4 basic statements ready for refutation.

    (1) There are no residential quarters (with latrines, hearths, water pipes, sewage pipes, streets etc.) built on top of Londinium’s dark-earth-covered ruins of the 3rd century before the 10th century.

    (2) There are no residential quarters (with latrines, hearths, water pipes, sewage pipes, streets etc.) built below 8th-10th century Lundenwic in the time from 1 to 700 AD.

    (3) Lundenwic’s dark earth layers of the early 10th century sit in the same stratigraphic plane as Londinium’s dark earth layers of the 3rd century.

    (4) 8th-10th century Lundenwic is the missing western vicus of Londinium. Thus, Roman Londinium does not belong to the 1st-3rd century but to the 8th-10th century period, too.

    I could add:

    There are no residential quarters (with latrines, hearths, water pipes, sewage pipes, streets etc.) built on top of the dark-earth-covered 3rd century of the Roman Forum before the 10th century. Therfefore, Rome’s imperial Forum-complex does not belong to the 1st-3rd century but to the 8th-10th century period.

    This statement about the city of Rome also applies to the whole of Roman civilization and the history of the rest of the world.

    Gunnar Heinsohn

  7. Pingback: Romans, Christians and Gods | Louis Hissink's Crazy World

Comments are closed.