Phase I Findings
Phase I Findings:
The Barrymore mausoleum was constructed to house the remains of James Barry, the 4th Earl of Barrymore.[i] James Barry’s body was placed in the crypt some time after his death in 1747 or 1748; at the time of his death he was in his late 70s or 80s (sources vary). Barrymore by Rev. E. Barry[ii] states that his memorial plaque was not completed until 1753; if that is the year that the crypt was opened, James’ remains must have been interred elsewhere until then. Later, the remains of other Barry relatives were placed in the crypt.
The crypt was vandalized more than once, the first time in1894 or earlier and more recently in the past 10-20 years. [iii] According to a family member, in the 1940s some of the remains were removed to prevent further desecration and reburied elsewhere. No records have been found to indicate which remains were removed or where they have been reburied. The plaque honoring James Barry is still in the mausoleum. No other memorial to him has been identified, nor any records to indicate whether his remains were among those that were relocated.
When the crypt was opened in August 2015 two coffins were discovered, along with a plate indicating that an urn containing the ashes of a relative, Charlotte Smith-Barry, née Cole (1847-1933), had once been there. Two sets of remains were identified in the coffins and designated Barrymore1 and Barrymore2.
Forensic Examination
The remains were incomplete and had been mixed in the coffins. The skull found in the second coffin appeared to fit a mandible and first cervical vertebra of the skeleton in the first. However the possibility that it belonged to the skeleton in the second coffin could not be excluded because those remains were missing some cervical vertebrae. A skull found in the first coffin did not fit that skeleton and showed evidence of previously having been buried in the ground. Anecdotal reports indicate that it may have been taken from remains buried elsewhere and erroneously placed in the crypt. No similar evidence was found on other parts of either set of remains in the crypt. Because of the missing vertebrae in the second coffin no conclusion could be drawn about whether this skull fit with those remains. The anthropometric points and all sutures for this skull were obliterated and there was extensive coverage of the scalp by dried and mummified skin, thus obscuring important landmarks for analysis of age and gender. Two left humeri were found in the second coffin; one fit with the left ulna in the first. Both sets of remains showed arthritis and other age-related pathologies.
The individual in the first coffin (Barrymore1) was evaluated as an elderly male. The pelvis of the individual in the second coffin (Barrymore2) showed a mix of traits, which precluded identification of gender. No conclusive forensic evidence was obtained to identify the two individuals. There were no coffin plates or other identification markers.
The estimated ages at death of both individuals are consistent with that of James Barry but not with those of the sons who succeeded him. One son, John Smith-Barry, died at the age of 59. A brother, David John Barry, predeceased James and probably was in his 60s when he died. There is insufficient information on the husbands of James Barry’s female descendants to determine whether any fit with the estimated ages at death of Barrymore1 or Barrymore2. The skull possibly associated with the individual in the first coffin, Barrymore1, showed a prominent nose; portraits of James Barry also show a prominent nose.
Five bone samples were taken, three from Barrymore1 for DNA testing and possible radiocarbon dating, and two from Barrymore2 for gender identification and DNA testing.
DNA Test Results
DNA testing conducted at Family Tree DNA indicated that Barrymore1 was a male and the test results are consistent with the history of the Barry family. DNA testing identified values for 14 YDNA STR (short tandem repeat) markers, confirming his gender. The results show that he was a member of haplogroup (deep ancestral group) R1b, the most common in Europe.[iv] About 80% of the more than 120 men with the Barry surname who have done YDNA tests are in that haplogroup.[v]
The YDNA results show possible distant paternal relationships between Barrymore1 and three groups of men in the Barry DNA project who are in haplogroup R1b. (Table 1) These groups have similar 12-marker haplotypes (STR patterns), but diverge at 25 markers. Twenty-five marker results are not available for Barrymore1 so it is not possible to determine the probability of relationships at that level.
The first is a cluster of about 30 men, the largest in the project, in haplogroup, R1b-U152-Z49. R1b-U152 is found in significant numbers in the test results for men with ancestry in Flanders, the probable place of origin of the Barry family. The men in this cluster had a common ancestor in Ireland around the 12th century, when the paternal ancestors of the Earls of Barrymore first came to Ireland, during the Cambro-Norman invasion. Ancestors of several men in the Z49 group lived near Barry strongholds in County Cork but none have records showing a relationship to the Earls of Barrymore.
Table1: Comparison of 12 Marker STR Test Result
(Non-matching markers in red)
Comparing the Barrymore1 results to the modal values for the Z49 group indicates that the probability of a common ancestor ranges from about 20% over 24 generations (600-720 years before the present) to nearly 50% over 36 generations (900-1080 years). If the Barrymore1 sample is compared to the minimum to maximum range of values for the Z49 group the probabilities are significantly higher. See Table 2. The mismatches between Barrymore1 and the Z49 group modal are from three rare values in his results. These may be unique to Barrymore1’s recent paternal line; thus the relationship may actually be closer.
SNP (single nucleotide polymorphism) tests of Barrymore1 for haplogroup U152 and Z49 were unsuccessful so the relationship remains unconfirmed.
There is also a potential match to a group of 11 men in haplogroup R1b-DF23. One of these men comes from a family that has a strong tradition of a relationship to the Earls of Barrymore but to date no documentation of this has been found. All of his current YDNA matches are to men with the Berry surname variant who have English ancestry. Further research and DNA testing are required to determine his relationship to the Anglo-Norman Barry family.
A third possible relationship is to a smaller group of nine men in haplogroup R1b-L21-L159.2 whose ancestors were from County Limerick. A comparison of Barrymore1’s results with the modal values for this group yields results identical to those for the Z49 group, because their 12 marker modal values are the same. However, unlike the Z49 group, the probability does not increase when the maximum to minimum ranges are taken into account, since that range does not encompass any of the unusual values shown in the Barrymore1 results. (Table 2)
There is a tradition in County Limerick that some Barry families were descendants of a younger son of a Viscount Buttevant, who was of the same paternal ancestry as the Earls of Barrymore; however, there are no records to confirm this belief.[vii]
A standard work on Irish surnames claims that most Barrys in that region were from an Irish clan, O’Beargha.[viii] R1b-L21-L159.2 is most common in the area around the Irish Sea, and has not been found among descendants of Norman or Flemish families. Thus this cluster of Barry project members probably was of Irish decent, rather than Cambro-Norman or Flemish, and less likely to be related to the Earls of Barrymore.
Table 2: Probability of Common Ancestry between Barrymore1 and Barry DNA Project Groups (12 Markers)
None of the four members of the Barry project whose family histories show a documented relationship to the Earls of Barrymore are genetically related to Barrymore1. [xi] They are either in a different major haplogroup (I1) or have so many mismatches with Barrymore1 (5 or more out of 12 markers) that the probability of a common ancestor in the past 900 years is less than 1%. There are also gaps and inconsistencies in their family records that raise questions about the integrity of their pedigrees.
Barrymore2 was also an older individual, whose gender could not be established from physical characteristics. DNA testing indicated that Barrymore2 was a male. Very few YDNA markers were recovered and his haplogroup could not be determined. The test results for Barrymore2 indicate a very low probability (less than one-tenth of one percent) of a paternal relationship to Barrymore1. (See Table 3.) He may have been the husband of a female Barry relative or there may have been a break in his paternal line such as an out of wedlock birth or undocumented adoption. Alternatively, Barrymore2’s test sample may have been contaminated by vandalism or degraded by age, or there may have been a testing error so that the results may be unreliable. Additional DNA testing of Barrymore2 is not feasible.
Conclusions
The results of forensic examination and DNA testing are consistent with the hypothesis that Barrymore1 was James Barry and that he shared a common ancestor with the group of men in haplogroup R1b-U152-Z49. The evidence to date, however, is insufficient to establish this hypothesis to a rigorous statistical standard of at least 95% confidence. Additional evidence, both forensic and genetic, may be able confirm the identity of Barrymore1 and his relationship, if any, to living individuals. Barrymore2 was unrelated to Barrymore1 and to any living men with the Barry surname who have done YDNA testing. His test sample is of poor quality and possibly contaminated. It is inadequate for further testing.
Table 3: YDNA Test Results for Barrymore Samples
(Family Tree DNA 12 Marker Panel)
Tentative Plans for Phase II:
The following are possible options for a second phase of the project:
· Continue efforts to identify Barrymore1 and assess relationships to living individuals. These could include:
o Re-examine the remains for further evidence of identity.
o Conduct radiocarbon dating for comparison with James Barry’s dates of birth and death.
o Attempt facial reconstruction of the skulls for comparison with portraits of James Barry.
o Take additional samples and conduct testing on Barrymore1 at another laboratory to include full YDNA sequencing, or selected SNP tests if full sequencing is impractical or too costly.
· Identify other remains of deceased members of the Barry family for forensic and genetic analysis.
· Test other men from the Barry family, especially those likely to be related to the Earls of Barrymore, and conduct historical research on their ancestry.
[i] Historical records indicate that James Barry was a direct descendent of the Barry family that came to Ireland in the 12th century during the Cambro-Norman invasion. Studies indicate that in similar families with well-documented lineages, some 1-2% of births involve an incorrectly attributed paternity. (See Appendix C.) That suggests a 15-30% probability that there was some discontinuity in the line of descent from the early Cambro-Norman Barrys to James. For this reason, additional testing of other documented descendants of that family will be required for comparison with the results presented here.
[ii] Published by Guy and Company, Cork, 1902. p. 122
[iii] A fragment of newspaper dated 1894 was found in one of the coffins in the crypt during a forensic examination. Recent graffiti was also found on the walls.
[iv] Estimated using Athey (http://stevemorse.org/dna/hapest.php) and Nevgen (http://www.nevgen.org) haplogroup predictors with 12 markers. The first 12 markers were used for this estimate because the two other markers tested were in higher-level panels and the other markers in these panels were not available to provide context for the calculations. Adding those markers to the Athey calculator returns a value of 85% for R1b; the Nevgen calculator does not yield a coherent value.
[v] See the results from the Barry Surname Project at Family Tree DNA at https://www.familytreedna.com/groups/barry/about/background
[vi] This table shows a total genetic distance of four between the DF23 group and Barrymore1; however, two of the differences are on polynomic markers (those with multiple values) and according to the “infinite alleles” method of determining genetic distance these are counted as a single mismatch.
[vii] James Grene Barry, “Ancient Mural Inscriptions, County Limerick” in The Proceedings and Papers of the Royal Society of Antiquaries of Ireland, (Dublin, University Press: 1890) p. 50
[viii] See Rev. Patrick Wolfe, Irish Names and Surnames (1913): http://www.libraryireland.com/names/ob/o-beargha.php
[ix] Genetic distance indicates the number of non-matching markers
[x] 24-generation estimate from Family Tree DNA Tip Calculator based on individual mutation rates for each marker. 30- and 36-generation values from McGee Calculator (http://www.mymcgee.com/tools/yutility.html?mode=ftdna_mode) which uses average mutation rates
[xi] The family histories are based on a privately published book, Barry Lore, https://dcms.lds.org/delivery/DeliveryManagerServlet?dps_pid=IE745209. Barry Lore bases its pedigree information on a family bible as well as on two other private works, The Records of the Barry Family, Volume I, which is available online on Ancestry.com and History of the Barrys, available at the New England Historic Genealogical Society. The pedigrees differ in several places from those in Barrymore and in Burke’s Irish Family Records. They also include middle names for some individuals derived from the names of related families; however, these names do not appear in Barrymore or Burkes, or in other Irish records of the Barry family. In one case, the claimed relationship to the titled members of the Barry family may be based on family lore rather than documentation and in another there is a possible discrepancy between the date of death of a Barry ancestor and that of his son’s birth.
[xii] The eight STR values available for Barrymore2 are insufficient to establish his
haplogroup. They appear to relate most closely to three haplogroups, O, Q and T, which are rare in Europe suggesting that there may be errors in the derived values.