The House of Tyndall
The House of Tyndall
Confide non confundar
A descent in the male line, 1010–2022
The House of Tyndall
Projects · DNA

The descent of a
chromosome

The documents on this site carry the male line from 1010. The Y-chromosome carries it from the Bronze Age. This page explains what was tested, how to read the result, what it says about where the family was before the records begin, and how any Tyndall can add his own line to it.

R‑Z40708where the line arrives
In plain words

Why a chromosome can be read like a surname

A man carries one Y-chromosome, and he got it from his father, who got it from his. It does not mix with the mother’s side the way the rest of the genome does, so it passes down the male line almost unchanged — the same path a surname takes. Over the centuries it picks up small copying errors, mutations, at a roughly known rate, and once a mutation appears every male descendant carries it. That gives two things at once: a signature that says which male line a man belongs to, and a clock that says how long ago two men’s lines parted.

So one living man’s test is a record of every father above him. Read upward, it names the branch his line sits on and the branches it shares with other tested men, back to the first farmers. Read against burials whose DNA has been sequenced, it puts his ancestors in a population, at a date, in a place.

The paper trail and the chromosome measure different things and barely overlap. The documents end at Siward in 1010 and are silent above him. The chromosome is coarse for the last thousand years and sharp for the four thousand before. Where they meet, in the late thirteenth century in Tynedale, they agree.

Haplogroup the branch
A named branch of the human paternal tree, defined by a mutation every man on it shares. R-Z40708 is the twig this family sits on; R-P312 is the great bough it grows from.
SNP the mutation that names a branch
A single changed letter in the chromosome. Each SNP marks one branching point; the names (Z40708, DF19) are catalogue numbers.
STR the quick match
Short repeated sequences that vary faster than SNPs. Cheap to test, good for finding living relatives, too coarse to date deep branches. The 12-, 37-, 67- and 111-marker tests are STR tests.
Terminal branch
The youngest named branch a tester belongs to. Everyone on it shares a common ancestor more recent than everyone off it.
Private variant
A mutation found so far in only one tested man. It becomes a branch the day a second man is found who shares it.
TMRCA
The estimated date of the most recent common ancestor of a branch: when the men on it last shared a father. Always a range, never a year.
What was done

Three tests, two laboratories, one line

TestWhoWhat it is forStatus
Big Y-700
FamilyTreeDNA		Waltheof IIIReads millions of positions on the Y-chromosome and places the line on the paternal tree to its terminal branch. The result every figure on this page rests on.Complete
Big Y-700
FamilyTreeDNA		Siward IIA second sample one generation up. Shows which private variants are shared and closes the branch to two men; the first step toward a modern Tyndall group on the tree.In the laboratory
Y-37 · Y-67 · Y-111
FamilyTreeDNA		Waltheof IIISTR panels, for finding living matches and comparing against surname projects.Complete
Autosomal
Ancestry		Waltheof IIIThe whole genome, all lines at once. Finds the recent heartland rather than the deep origin; here it returns the Nene valley of Northamptonshire.Complete
Ancient connections
FTDNA Discover · YFull		—Comparison against sequenced burials and against the independent YFull tree, which dates the same branches from a different sample set.Ongoing
What it says

Four thousand six hundred years, in four chapters

The result in one line: the family’s male line is R-Z40708, a branch that coalesces in the late thirteenth century in Tynedale and is shared today with the Tinlins and Tinnings of the border country. Above that, the chromosome tells a story the documents cannot.

2600 – 1050 BCE
The Dutch coast
The line arrives with the Bell Beaker people on the Lower Rhine and stays. Men in the ground at Oostwoud, Ottoland and Zwaagdijk share it. For fifteen hundred years and six consecutive branching points every match is in the Low Countries. About 1450 BCE the kindred divides three ways: one road up the Elbe into Saxony, one with the Langobards to Pannonia, and this one.
c. 1050 – 950 BCE
North, across the water
At R-Z21280 the Low Countries signal stops. At the next branch, R-Z21279, the first Scandinavian matches appear and no Dutch or Belgian ones follow. A population that moved looks exactly like this in a phylogeny. From here the evidence points to Sweden.
950 BCE – 1288 CE
Sweden, and the silence
One more Swedish match at R-Z40722, about 550 BCE, and the line’s only other surviving branch parts there. Then eighteen hundred years in which no branch forms and nothing is recovered from the ground: the Roman Iron Age, the Migration Period, the Anglo-Saxon settlement and the whole of the Viking Age pass without a trace but the line’s own continuation. Where in that span it crossed to Britain the chromosome does not say; the documents put Siward, a Dane in Cnut’s service, at York by 1033.
1288 CE – today
Tynedale
The terminal branch coalesces in the valley the surname comes from, in the same window as Richard and Robert at Tansor. Twelve men descend from it; two of them, Tinlin and Tinning, are still in the border country under worn forms of the same name. Below them, seven centuries and four mutations: this line, and nobody else.
How to read the chart below
Scroll, and the map follows the line down from the Lower Rhine to Tynedale. Each stop is a branching point on the paternal tree: its name, its date, and how many tested men alive today descend from it. The marked burials are men in the ground who share the line at that point. Cousin branches that left the chart — the Visigoth at Girona, the Langobard in Pannonia — are shown leaving.
2600 BCE
the Lower Rhine
R-DF19
4,000
men alive today who descend from this point
The same descent, counted

From four thousand men to twelve

Each bar is the number of tested men descending from that branching point. The line falls by two orders of magnitude and never recovers.

Every terminal branch narrows — that is what a terminal branch is. What is unusual here is that the narrowing runs continuously across four thousand years, and that the line survives it rather than ending.

The shape of it

Four numbers that describe this line

45matches at twelve markers, worldwide
3living men tested at the terminal branch
1,838years with no surviving branch
4,600years traceable to Bell Beaker burials

The forty-five is the one that matters. Twelve markers is the lowest resolution the test offers — designed to catch distant and even accidental matches, and it does not require anyone to have upgraded. Testers in Northwest European haplogroups routinely return several hundred to several thousand at that level. One published example runs to 16,978.

Forty-five, worldwide, is not a sampling gap. It is the line itself.

The anomalies

What is unusual about this line, stated as facts

Eighteen hundred and thirty-eight years without a surviving branch.
Most lines in R-P312 branch every few centuries. This one forms no branch anyone has tested between about 550 BCE and 1288 CE.
A line kept to one heir at a time, by custom or by estate; or branches that formed and died; or cousins alive and untested. The last is weakest, because the populations this line runs through are the most heavily tested on earth.
Forty-five matches at twelve markers, worldwide.
Twelve markers is the coarsest test offered and routinely returns hundreds or thousands of matches for a Northwest European haplogroup.
The line is small and old, and has been for a long time. The number is the line itself, not a sampling gap.
Two mutation speeds in one chromosome.
Seventy-seven years per mutation from 550 BCE to 1288; a hundred and eighty-five from 1288 to the present.
Late fatherhood among landed heirs, which the pedigree shows directly: two generations fathered by men in their late forties. Four events is a small sample.
The Tindall surname project does not match us.
The project gathers men named Tindall, Tindal, Tyndall and Tyndale from across the world. None of them share this line.
The surname is locative, from the valley, and was taken independently by unrelated families who lived there. The line this site follows is one of several that carry the name, and it is the one that carries the documents.
The closest ancient autosomal genomes are eleventh-century Swedes.
For a family documented in England from 1010 and in Northamptonshire from 1247, the nearest whole-genome matches in the ground are Scandinavian.
A Scandinavian house at the head of the line, which is what the documents say of Siward.

On the central one

Most male lines die out — that is the ordinary outcome, and branching-process mathematics predicts it. A few expand explosively; those are the famous cases, and they are broad and shallow. This is the third thing: a line that neither died nor spread, for sixty generations.

The last fork on the way down is at R‑Z40722, about 550 BCE. Everything else under that node is one other branch — an Englishman and a Swede whose own common ancestor lived about 880. Between the fork and 1288 this line forms no branch that anyone has tested.

Why undertesting is a weak explanation here
The usual objection to a long unbranched block is that the cousins exist but have not tested. That is strongest in undersampled populations. This line runs through R‑P312 — the most heavily tested branch of the human paternal tree — in England, Scandinavia and the United States. A gap that survives that testing density is a different object from a gap in a thinly sampled clade.
And the deep bottlenecks in the literature are the wrong shape
Long unbranched stretches do appear in published work, but overwhelmingly in hunter-gatherer lines, island populations and sequestered groups, and overwhelmingly before 1000 BCE. This one runs through the Roman period, the Migration Period, the Anglo-Saxon settlement and the Viking Age — on the European mainland, in the middle of everything.
Mutation rate

The same line runs at two different speeds

Mutations accumulate per generation, not per year. The two halves of this descent differ by a factor of about two and a half — and the change falls where the family becomes a landed English one.

550 BCE to 1288 CE
77
years per mutation
About 24 variants across roughly 1,838 years. Short generations, young fathers.
1288 CE to the present
185
years per mutation
Four private variants across roughly 740 years. Long generations, late heirs.

A coverage or calling problem cannot produce this. One sample, one sequencing run, one callable region — an artefact would depress both halves. The ancient leg sits where the published rate constants predict. The recent leg does not.

The pedigree offers a mechanism. Landed inheritance produces late marriage: heirs wait for the estate. Two generations here were fathered by men in their late forties — John of Deene fathering Sir William at about forty-nine, and Sir William Tyndall KB fathering Sir John at about forty-five.

Helgason et al., Nature Genetics 47 (2015); Adamov et al. (2015); Karmin et al., Human Genetics (2017). On paternal age, Kong et al., Nature 488 (2012). Four events is a small sample and the recent figure carries wide uncertainty. Node ages on this page are FamilyTreeDNA Discover estimates; YFull, on an independent sample set, puts R‑Z21280 at about 1174 BCE and R‑Z21279 at about 1074 BCE, a century older and within the published intervals.

Ancient connections

Men in the ground who share this paternal line

Each is a sequenced burial sharing a common paternal ancestor with this line at a named node. They do not prove descent from the individual — they place the line's ancestors in a population, at a date, in a place.

BurialDateShared ancestorRarity
De Tuithoorn, Oostwoud
Bell Beaker, Netherlands		2579–2211 BCER-DF19—
Kromme Elleboog, Ottoland
Bronze Age, Netherlands		2500–2100 BCER-DF88—
Zwaagdijk, Wervershoof
Bronze Age — bronze sword, Baltic amber		1620–1311 BCER-S4281—
Valkenburg 107763, Zuid-Holland
Roman-period Netherlands		c. 179 CEupstream—
Hiddestorf 1 and 2, Lower Saxony
Saxon — Gretzinger et al. 2024		300–500 CER-Z171121 in 484
Otto the Great and Henry II
Liudolfing emperors, Magdeburg and Bamberg — Ringbauer et al. 2026		912–1024 CER-Z171121 in 484
Szeleste 437-2 and 599, Vas
Langobard, Pannonia		475–575 CER-FT3541491 in 1,900
Koksijde 17, Flanders
Merovingian coast		c. 619 CEupstream—
Plaza de l'Horta, Girona
Visigothic, Catalonia		400–700 CER-Z171121 in 484
Öland 1028
Viking Age, Sweden — the one Scandinavian burial		772–1014 CER-Y6237, two nodes above Z21280—
Akbari et al. match
the closest ancient connection held		c. 1050 BCEZ21280 era1 in 13,000
What the cousin burials do and do not tell you
The Langobard, the Visigoth, the Saxons of Hiddestorf and the Ottonian emperors all split from this line around 1450 BCE, and what they did afterwards is their story rather than this one. What they establish is the character of the population this line sat in — substantial and continuous enough to seed the peoples who broke up the Roman west, and to put one of its branches on the imperial throne. From the 1450 BCE node the kindred took three roads: up the Elbe into Saxony, which produced the Liudolfings; south with the Langobards into Pannonia; and north into Sweden, which produced this line.

Rarity figures are FamilyTreeDNA Discover outputs, computed against its tested-customer base rather than the world population.

The living

Who else is out there

MatchWhere the paper trail leadsStanding
TinlinScotland, the border countryTerminal branch, R-Z40708
TinningCumberland — earliest ancestor James Tinning, b. c. 1646Terminal branch, R-Z40708
FreesePeter Frese, b. Lübeck, d. Viborg 1652Last surviving European match at 37 markers
Hudson / HutsonStaffordshire, and Henrico County, VirginiaDeep cousins clearing the threshold at P312 depth
SöderlinGodegård, Östergötland, SwedenExact match at twelve markers; R-Z17125, confirmed within DF19

Tinlin and Tinning are the same name. It is locative — from Tynedale, the valley of the Tyne — and it wore down where the family lost the land that anchored it. One of these men traced his own name backward, independently, and found it running toward the older form.

Beyond them there is nothing at 67 or 111 markers in England or America. The match list is bimodal: two close border-country cousins, a scatter of very deep ones, and an absence in between where a normal line would have its relatives.

Where the two records meet

The paper trail was built first, and the testing did not break it

The documentary reconstruction was made from records before any sample was taken. It was therefore falsifiable, and the genetics could have contradicted it at several points.

What the documents saidWhat the testing returned
A Northumbrian family, from TynedaleTerminal branch matches in the Northumberland and Cumberland border country, carrying the same surname stem
The English pedigree opens in 1288, with Robert Tyndall of TansorThe terminal branch coalesces in the late thirteenth century
Tansor, in the Nene valley, held from 1247Autosomal testing returns a concentrated English signal on Thrapston, Raunds, Denford, Ringstead and Oundle
Entry through the Anglo-Danish nobility rather than mass settlementA narrow, non-broadening line — not the bushy profile a settler population leaves
Siward, a Scandinavian magnate, at the head of the documented lineSwedish matches at every upstream node from R‑Z21280 down, and no Low Countries match after it; the closest ancient autosomal genomes are eleventh-century Swedes
Two instruments, two ends of the same chain
Y-DNA reads the unbroken male line and lands on the deep origin. Autosomal reads the whole genome and lands on the recent heartland. Neither could produce the other's answer — and both agree with a reconstruction built before either was run.
Are you a Tyndall?

Every man who carries the name can add a line to this tree.

This is an awareness project as much as a family one. Each new test either places a man on this branch and dates where his line parted from ours, or shows that his Tyndalls are a different family that took the same valley’s name. Both answers are worth having, and both go on the record here.

Who we are looking for
Any man who carries Tyndall, Tindall, Tindal, Tyndale, Tindale, Tinlin or Tinning, or who descends in the male line from one who did, anywhere in the world. Women can take part through a brother, father or uncle.
What to order
The Big Y-700 at FamilyTreeDNA is the test that places a line on the tree and dates it. A Y-37 is enough to say yes or no, and can be upgraded later without a new sample. Join the Tindall surname project when you order; the comparison is automatic.
What it costs and how long it takes
A Y-37 is in the low hundreds of dollars and returns in about a month; a Big Y is several times that and takes two to three. Both go on sale several times a year. If cost is the only thing stopping you, write to us.
What happens with the result
We place it on the tree, write it up on this page and in a project, and tell you what it means for your own line. Your name appears only if you want it to. The raw data stays yours.
Write to us about testing →

Testing through FamilyTreeDNA, Houston, Texas. This site has no commercial relationship with any testing company.

Honestly

What the genetics cannot do

It cannot identify a person. No sample exists from Siward, from Waltheof, from Uhtred, or from any Tyndall of Hockwold. Descent from named individuals rests on the documents.

It cannot say why the line stayed narrow. A line kept narrow by inheritance and one kept narrow by chance produce the same signature. For the documented centuries the mechanism is visible in the records. For the eighteen hundred years before, it is not.

The dates are estimates with wide margins. The terminal coalescence carries a 95 % interval running from the eleventh century to the sixteenth. It is a distribution, not a year.

And parts of it are thin. Three tested men at the terminal branch. Four private variants. Single modern matches carrying geographic weight.

What would sharpen it
A second sample on this line is in the laboratory: a father’s Big Y, which will show the four private variants shared and close the branch to two men. Beyond that, a collateral descended from a brother somewhere in the English or Virginia generations, to date an intermediate node and place the line on an independent tree with its own age model.
Sources and further reading

Where the figures come from