Among the historical family archives in Trinity’s Library are the papers of the Clements family, of Ashfield, Co Cavan, and Lough Rynn, Co Leitrim. This year, over a century and a half after it was written, an important missing element of the archives was presented to the Library; it is a travel journal written in the 1840s by Henry Theophilus Clements. However, this is not the full story. Apart from the journal, the donor Saoirse Fitzallen also presented her transcription of the diary. This makes the diary significantly more likely to be used – simply by making reading and searching it easier – and may have the effect of drawing scholarly attention to the main Clements family collection.
Nathaniel Clements (d. 1771) was the 18th-century equivalent of a Minister for Finance and a very influential political figure. He became MP for Duleek in 1727. It was he who built the house now know as Áras an Uachtaráin, the home of the President of Ireland. His grandson Henry Theophilus Clements (24 July 1820‐7 January 1904) was the son of an Henry John Clements (1781-1843) MP for Leitrim, and for Cavan. He was the nephew and inheritor of the estates of the ‘wicked’ Third Earl of Leitrim; and he was a soldier who spent most of his childhood in France.
The Clements family papers have been in the Library since the mid-1970s. The collection contains the usual range of financial and estate-management material, and is particularly rich in travel journals. This is why it is so gratifying to be able to return this missing item to the collection and the story of its discovery is interesting. Our guest author, and the donor of the journal, Saoirse Fitzallen, writes:
“I found the journal while on a wee stroll to stretch my legs on holiday in Norfolk. It had been left outside a cottage with several other journals and other bric- a -brac on a ‘help yourself’ shelf – they had been cleared out of a holiday home that I later discovered had belonged to one of Henry Clements’ descendents. Researching into the author’s life, and transcribing the journal, became an incredibly rewarding project for me during the first UK Covid lockdown in 2020. I figured out who the author was once I’d deciphered his name and looked up where Ashfield was. I gradually made the connections by using Debretts and other Google searches and reaching out to Irish historians who put me in touch with surviving members of the family all of whom were very helpful.
This journal chiefly charts the journeys taken by Henry, his mother, and his sisters to cheer them all up after the death of his father Henry Clements senior in 1843. This alone would have given me a soft spot for him but I am so grateful that he took the time to write his adventures down and that in transcribing his journal I was able to learn so much about his life and the lives around him.
The journal is an extended progress through Europe with Henry, his family and staff staying in grand hotels and beautiful houses, seeing amazing artefacts, marvelling at feats of engineering and architecture, attending operas, plays and entertainments while eating and drinking in all the best spots with all the best people.
At the back is a proto Tripadvisor which rates all the hotels in which he stayed. If you, as I did, follow his journey using Google Maps, you can see most of what he saw and might be tempted, as I was, to follow in his footsteps. The journey won’t be so long – one of his day-long trips would take us just half an hour in a car. Some of the treasures will have been lost or locked away but we’d still be able to see that double-helix staircase in France. Horses aren’t used to carry stones to the roofs of buildings anymore so that’s a sight we’ll never be afforded but I wonder if there’s still a transparent alabaster window in that monastery in Italy? I wonder if a rainbow is still created as the falling water hits those rocks in Switzerland? What has replaced the ladder they used to descend to that Spa?
I am so very proud to be able to donate this journal and my transcription of it to Trinity College to join the rest of the Clements family papers and hope you enjoy reading it as much as I enjoyed working on it.” Saoirse Fitzallen
TCD MS 503 is principally comprised of John of Worcester’s ‘Chronicula’, a text which describes itself as having been derived from the more celebrated and much more elaborate Worcester Chronica chronicarum, a twelfth-century chronicle, the authorship of which has been attributed variously to Florence and/or John, both Worcester monks. As the diminutive form of its Latin title suggests, the Chronicula itself is of diminutive size, perhaps designed to be easily portable. Its text has never been printed before and those interested in its contents have previously been obliged either to order a microfilm copy or to visit the archives attached to the magnificent Long Room of Trinity College, Dublin (not a hardship!). The manuscript has now been digitised and made available online https://doi.org/10.48495/000004234. Sixteenth-century annotations in the manuscript, as shown in the image below, record that in 1573 the manuscript was sold to the antiquary William Lambard and then, in 1594, Lambard gave the manuscript to another English antiquary, William Howard. Both men were connected to a circle of antiquaries who had various interests in the literature of the Anglo-Saxon and Anglo-Norman periods, including such figures as William Camden, Laurence Nowell, Henry Spelman and Robert Cotton. William Howard had, in 1592, published an edition of the Worcester Chronica chronicarum,which presumably indicates why he was also interested in TCD MS 503.
Although we do not know the exact date of the Chronicula’s composition, the late 1130s has recently been suggested. For the most part the work is derivative of the main Chronica chronicarum, a text which, itself depending on a range of sources, gives in annalistic form a history of the world to the year 1140. In deriving its information from the Chronica chronicarum, the Chronicula heavily summarises and abbreviates events and omits much detail. The Chronica chronicarum is set out in annalistic form, whereas the Chronicula is arranged according to the reigns of the Roman emperors and the year in which they came to power (reckoned according to a system developed by Marianus Scotus in his Chronicon who sought to re-date the Dionysian system by 22 years, a chronicle which itself provides an important model for the Chronica chronicarum). Information that can be found in separate annals in the Chronica chronicarum can therefore seem lumped together in the Chronicula.
It is probably the derivative and abbreviated nature of the Chronicula that means it has so far attracted relatively little attention. But there are important reasons why it deserves further notice. The first is that it seems to have been written, for the most part, in the hand of John of Worcester himself. Given John’s authorial role in the writing of at least parts of the Chronica chronicarum, we are therefore being given access to an associated work, a text that is less formal than, but nevertheless complementary to, the main chronicle. A second is that the Chronicula is not entirely derivative in form and actually contains unique information. Its main text opens with a geographical description of Britain, not found in the Chronica chronicarum and which is closely connected to a similar passage in the F manuscript of the Anglo-Saxon Chronicle. It also invokes the work of Hugh of Fleury to a greater extent than the Chronica chronicarum, and, perhaps most strikingly of all, contains a range of poems that cannot be found elsewhere. Three poems in particular stand out: those that celebrate the deaths of Edward the Confessor in 1066, of Harold, son of Godwine, also in 1066, and of Bishop Wulfstan II in 1095, the very man who was credited in Worcester texts with initiating various historical projects. The poems are not accomplished in literary terms. But they are striking for their wistful attention to the Anglo-Saxon past and for a degree of anti-Norman sentiment. If they were composed by John himself, some 70 years after the Norman Conquest, they provide important evidence for the sort of nostalgia for the past and the sort of anti-Norman hostility that could continue to be voiced by at least one Worcester inhabitant. We are left to wonder why they were not included in the main Chronica chronicarum.
For whatever reason, John ceased writing his Chronicula once he reached the annal for 1123, on line 23 of fol. 113v. A new hand (one of four in total that have been identified as working on TCD MS 503) continues the story, copying annals down to 1141. Because the annals from 1123 to 1141 have material with a clear Gloucester focus, it is possible that the Chronicula had travelled to Gloucester where it was then updated. John’s motivations for writing the Chronicula remain elusive. At one point in the manuscript he says that anyone wanting to find out more details about a certain historical episode ‘will find them written in chronological order and more fully in the chronicle of chronicles [i.e. the Chronica chronicarum]’, suggesting that he anticipated that the Chronicula would reach at least some kind of audience. Was this an audience within Worcester itself? Or, given the later connection with Gloucester, was it designed to be sent there? We will probably never know John’s precise motivations, but the Chronicula is important as an extra example of this twelfth-century historian’s working methods. Scholars are at present re-doubling their efforts in connection with Worcester historiography. An edited volume on the cartularies and chronicles (and other texts) produced at Worcester in this period is forthcoming in May 2022, entitled Constructing History Across the Norman Conquest: Worcester, c.1050-c.1150. And the Chronicula itself is actively being edited for the Oxford Medieval Texts series. In the meantime it is an enormous boon for scholars around the world to have access to these high-quality images of TCD MS 503 so that they can engage directly with John’s work for themselves.
D. A. Woodman
Robinson College, Cambridge
Virtual Trinity Library is a digitisation initiative of the Library of Trinity College Dublin’s most valued collections. It will conserve, catalogue, curate, digitise and research these unique collections of national importance, making them accessible to a global audience, from schoolchildren to scholars. https://www.tcd.ie/virtual-trinity-library/
By Dara Burke, architect, and Michael White, Northburgh Castle Conservation Group
The Northburgh Castle Conservation group in Donegal used 19th-century paintings held in the Library, as part of their Conservation Plan for the 14th-century Anglo-Norman ruin. The Plan is being launch by the American Ambassador to Ireland Claire D. Cronin, who has family connections to Donegal.
Northburgh Castle 3D model
Northburgh Castle, built in 1305, is an Anglo-Norman castle situated in Greencastle, north Donegal, and described in the Ulster Journal of Archaeology as “the largest and most impressive building of its kind in Ireland”. The castle was built by Richard Óg de Burgo, the 2nd Earl of Ulster, also known as the Red Earl. Northburgh is historically significant, marking the highpoint of Anglo Norman control of Ulster, and bearing witness to Edward the Bruce’s invasion of 1316, the collapse of the Ulster Earldom and subsequent Gaelic revival.
However, the castle has been neglected and the structure has deteriorated in the past 30 years. Because of this, in 2021 we set up the Northburgh Castle Conservation group and then commissioned a Conservation Plan for the castle, part funded by the Heritage Council. In parallel we also commissioned a 3D model of the castle from architect Dara Burke. Dara’s work was supported by some 19th-century paintings of the castle, done by Captain William Smith. These pictures, which are in the Library of Trinity College, the University of Dublin give a clearer picture of the structure of the castle before it became covered in vegetation in the 20th century. They were a constant source of inspiration and reference during the project.
We wanted the model to digitally record the site and the extent of the castle in 3 dimensions. The architect used a drone to scan the structure and capture close up details of the remaining towers at the site. The model was also used to generate scaled plans, sections and elevations of the site and the castle for use in the conservation plan.
3D view showing the model in point cloud model
Apart from recording details of the castle, we also wanted to use the 3D model as a way to promote awareness of the conservation project. We uploaded the final model to the website Sketchfab, where anyone can view and explore the models on any device.
Once the 3d scan was completed it was possible to create scaled two dimensional orthographic projections of the site. It was then possible to digitally overlay a survey carried out by archaeologist DM Waterman in the 1950s.
The original survey was published in the Ulster Journal of Archaeology issue 2 in 1958. It is very interesting to compare the current plan of the castle with the survey drawing as it was clear that the site had deteriorated and had become far more overgrown in the past 30 or 40 years.
And the accuracy of the survey which was carried out without any of the modern surveying equipment we have today is very impressive. By comparing the survey plan to the initial 3D scans, areas were identified which were not immediately obvious due to the trees and ivy that dominate the site in many areas. These areas were then scanned again to add more detail to the overall model.
3D scanning process
The architect Dara Burke created the digital 3d model of the site using drone-based, geo-referenced photogrammetry. Photogrammetry is a process that enables three dimensional measurements to be extracted from two dimensional data such as photos. We used a software application called Reality Capture along with a DJI Mavic drone to capture and organise the photographic images.
The Reality Capture software uses GPS coordinates from the drone flight to geo locate the model for mapping and to establish a rough location for each image recorded. It then analyses the data to detect the alignment and camera pose of each photo. By comparing the difference in parallax between points identified in two images, the software can accurately establish the three dimensional location of those points in space. This process is repeated millions of times to build up a full 3d model of the location being scanned.
The architect carried the scan over a period of 5 weeks during August and September 2021. He made multiple visits were made to the site and around 500-1000 drone photos were captured per visit. An initial scan was carried out to generate a low detail base model. Additional visits were then carried out to add detail in particular areas of interest.
Since the software captures colour information, it was important to capture the site in even lighting conditions. Overcast days are ideal for photogrammetry scanning as there are no hard shadows from the sun. By selecting days to scan with overcast weather and recording the data at broadly similar times of day it was possible to merge the datasets without issues caused by strong shadows or difference in lighting intensity. This method allowed far more detail to be captured than if the project was limited to one day of scanning.
The architect manually flew the drone in a series of orbits around the castle and aerial grid patterns of the overall site to capture a dataset of over 7000 images. Although it is possible to automate the drone flight for photogrammetry, manual control of the drone allows the operator to fly closer to the structure to capture more detail. Additional photos of the basement level and keep were captured using a digital SLR camera on a tripod.
After each scan visit the architect reviewed the new sets of photos then loaded them into the Reality Capture software to add detail to the model. Some sections of the scan such as the basement level were analysed independently and then merged into the main model. The software initially generates a sparse “point cloud” preview of the model which gives a sense of the overall form.
The point cloud is then processed into an 3d mesh with no colour information. The final step is to texture the mesh using the photographic data to generate the colour information for each point. All of this processing is extremely time consuming for the computer and took up 36 hours for the high detail meshing phase.
The model was processed in this way each time there was new data available, adding more detail to the 3d model each time. The final project file contained over 300 million points and consumed over 250gb of hard drive space.
The high detail 300 million point mesh was optimized down to between 300,000 to 3,000,000 points to make it viewable on the web using the Sketchfab online platform. In order to showcase the high level of detail captured, study models were created of smaller parts of the mesh. This allowed the south gate towers and basement level to be shown in high detail in a web browser.
Due to the amount of overgrowth and large trees on at the site some areas were difficult or impossible to scan effectively but the final scan delivered a good overall 3d model of the castle with high levels of detail. The southern gate towers were scanned and processed to a resolution of 1mm per point which is able to clearly show fine detail such as joints in the stone work and cracking around windows. The advantage of using drone based photogrammetry is that the inaccessible tall elements of the towers can be captured in great detail.
Trinity College, the Paintings of Captain William Smith and ‘Verified Views’
William Smith (1778-1832) was a captain in the British Royal Engineer’s Office, stationed in Buncrana, County Donegal. He was responsible for the building of a number of Napoleonic-era towers in Donegal, working under Major-General Charles Ross. He made many paintings of military installations, including a number of drawings and paintings of Northburgh castle in 1806, which he bequeathed to the Library of Trinity College, the University of Dublin. The entire collection is digitised and may be seen among the Library’s Digital Collections. There is also an online exhibition, curated by the Heritage Council’s architectural officer, Colm Murray.
The paintings by Captain Smith give a clearer picture of the structure of the castle before it became covered in vegetation in the 20th century. They were a constant source of inspiration and reference during the project. The paintings clearly show parts of the castle that are currently completely obscured by ivy and trees. While the 3D scan was being refined the paintings were referred to in order to see which parts of the site should be captured in further detail. Once the scan was completed we decided to see if some of the views captured by Captain Smith could be replicated using modern 3d software.
The architect imported the 3D scan into the software system Unreal Engine in order to replicate the views. Unreal Engine was originally developed for use in the computer gaming sector but has been adopted by many other industries in recent years as a powerful visualisation tool that can deal with high detail meshes in real time.
Once the model was set up in Unreal Engine, virtual cameras were created to replicate each of the views captured by Smith. In modern building planning applications a technique known as “verified view” is often used to communicate the visual impact of a new building proposal. This typically involves surveying a camera position in the real world and then creating a virtual camera with the same settings.
Although it is obviously is not possible to determine the exact position of Captain Smiths painting easel, the notes from the paintings made by Smith were used to establish where each painting was created. Together with historic maps and the 3d height data of the ground levels created by the 3D scan, the location for each painting was derived with a good level of accuracy
Cameras were placed at ground level and at eye height to replicate the view as seen by Smith. Once the cameras were in position the paintings were digitally overlaid and the camera positions were finely adjusted to match the paintings as closely as possible.
It was very impressive to see how three dimensionally accurate Smith’s paintings were when matched to an accurate 3d scan. By referring to Smiths paintings and comparing it to the matched 3D views it was also possible to identify clear architectural forms and details that are currently covered in a thick layer of overgrowth. It would be exciting to see the ivy removed from the towers to reveal the architectural features which are currently obscured but clearly visible in Smiths paintings.
To complete the study of the paintings an animated camera was then created that moved along a path pausing at each of the locations of Smiths paintings. The painting for each location was then composited over the 3D view to show how the castle looked from that location at the time the paintings were created in the early 19th century. This video gives a great sense of how dramatically the site has changed over the last 200 years and Smiths paintings are an invaluable historic record in this regard.
You can view the video here – https://www.youtube.com/watch?v=XCj335SpP80
With the model and Conservation Plan now complete in early 2022, the Conservation Group now plans to seek permission to make some physical interventions in an area of the castle where there is a risk of collapsing masonry. The Conservation Plan also suggested additional conservations steps to be taken in the next 2 to 3 years, starting with securing part of the structure of the South Tower Gate.
In parallel to this, we plan to ask Dara Burke to create an additional model of the castle as it appeared when first built in 1305. This will be based on a combination of the current 3D model plus the 1958 archaeological survey and information we have on comparable castles at Hearlech and Caernarfon Castles in Wales.
Lily (1866-1949) and Elizabeth Yeats (1868-1940), pictured above, originally moved to Dublin from London to join Evelyn Gleeson (1855-1944) in her newly established arts and crafts enterprise, Dun Emer Industries in 1902, where the printing of high-quality books and prints was overseen by Elizabeth and embroidery by Lily. The enterprise was named after the Irish mythological figure, Emer, who was renowned for her artistic and needlework skills, and Cúchulainn’s wife. However, after an acrimonious split with Gleeson, the sisters established Cuala Industries in 1908 taking their own areas of production with them. The ideology of both organisations was espoused in the original Dun Emer prospectus, which stated its desire to “make beautiful things” using honest and native materials in “the spirit and tradition of the country”. Both were female enterprises and almost exclusively employed and trained young women as assistants in arts and crafts
The Press, the dominant part of Cuala’s business, published handcrafted books by leading members of the Irish literary revival including Nobel prize-winning sibling William (1865- 1939), and prints designed by Irish artists, chief among them another sibling Jack Yeats (1871-1957). Lily’s embroidery department was also notable, but its output was smaller and its legacy harder to track as many of the domestic embroidered items, such as, clothing, tablecloths and bedspreads are no longer extant. Framed embroidered art works such as those in the National Gallery of Ireland and in private collections indicate the artistry and technical quality of the embroidered work of Lily and her assistants. Before moving to Dublin, Lily had established herself as a skilled artistic embroiderer working for six years in the late 1800s with May Morris, daughter of William Morris, in their world-renowned Arts and Crafts scheme.
Trinity Women Graduates (formerly Dublin University Women Graduates Association) has partnered with the Library of Trinity College to present an exhibition photographs, records and other historical documents from the Trinity Women Graduates’ archive and the College Records in the Long Room of the Old Library.