Hulley Family History >> Publications > PROPhile - Friends of PRO

pLongview Publishing
pNorth Cheshire Family Historian
pThe Manchester Genealogist
pCatholic Ancestor
pPROPhile - Friends of PRO
pCheshire Record Office
pOxford Dictionary of National Biography
pJournal of One-Name Studies
pMacclesfield Express
pMacclesfield Times

pGuest Book
pView or Sign

pContact Me
pSend Email


PROPhile - Friends of PRO

Transcription of the 1851 unfilmed census for Manchester and district

The 1851 census is a vital source of evidence for family and social historians, being one of the earliest data banks covering all persons on a single date. It is the first census that shows full details of family relationships, exact ages and occupations of all employed persons. Most importantly, it shows for the first time the birth town or village of all English-born people. This enables the researcher to go back into the pre-census period with a definite location for further work. Most of the returns have been microfilmed, surname and location indexed and analysed by researchers and family history societies. Their efforts have made the searching of the records a comparatively easy task and many of us are greatly indebted to them for their hard work.

In 1983, the Manchester and Lancashire Family History Society decided to surname index all the returns for the Manchester district and a programme of work was set up by John Coupe, the society's project officer. A team of volunteers was recruited and the first volume of surnames was published in March 1984.

The initial exercise immediately drew attention to a major problem with the Manchester returns. The microfilms used to index the details were, in many cases, extremely difficult to read, mainly because of water damage to the original returns which had been suffered whilst the returns were still under the jurisdiction of the Home Office. This water damage was reproduced on the film as large, ghost-like areas of white with the text being completely obliterated from the same area. Further enquiries at the Public Record Office revealed that in many instances the original returns were so badly damaged that it was not even worth filming them. Indeed, it was once believed that these returns had been completely destroyed, much to the disappointment of hundreds of family historians.

The work of indexing the filmed returns continued between 1983 and 1989 and it was decided to seek clarification from the PRO about the fate of the unfilmed returns. Although badly damaged - those affected were covered in a brown stain, caused by the water-based ink used by the enumerator having been washed all over the paper - they were all extant and stored with their companion returns in piece number order. A number of them were extremely fragile, some were in tatters and many appeared impossibly difficult to read. In late 1990, an analysis of all the unfilmed returns gave an indication of the extent of the problem which was much greater than at first envisaged.

A later comparison with the population tables of the district indicated that over 150,000 people were missing from the filmed returns and the areas affected included parts of Manchester - London Road, Deansgate, Hulme, Chorlton on Medlock and St George's sub-districts, as well as most of Salford - Greengate, Regent Road, Pendleton and Pendlebury sub-districts. Outlying areas also had large swathes of population unaccounted for - Ashton under Lyne, Oldham, Harpurhey, Blackley and Prestwich all had missing population amounting to over 50% of the total. This was a serious omission for historians and one that could not be replaced by other sources.

In early 1991, the Society sought volunteers from amongst their London-based members to transcribe the unfilmed returns and the work started in April at Chancery Lane. Susan Lumas of the Editorial Services Department was appointed as liaison officer between the PRO and the volunteers and kept an eagle eye on us. The work had to be carried on without the use of any form of artificial light and although an ultra violet light source was tried initially, the results were unsatisfactory, mainly because of the poor quality of the rather outdated apparatus available. Great reliance was placed on natural daylight and the teams successes in those days were almost entirely dependent on the vagaries of London's weather.

The returns were transcribed independently by two volunteers and their results were matched by a third, who arbitrated on any differences. A complete transcription of all readable text was made, including obvious errors, spelling mistakes and side notes. At one time, there were as many as fifteen volunteers working on the returns from a large table in the middle of the room either on a Wednesday or Thursday. Later, the work was transferred to a 'turret' room on the south western comer of the building, thus increasing significantly the amount of natural daylight available to the team who usually attended on one day each week, either Tuesdays or Thursdays.

After transcription the results were delivered to the Society's office in Manchester where another team of volunteers input the results into a computer prior to their preparation and eventual production on microfiche. The finished result contained a surname index, a location index and a full transcription of all recaptured data. A copy of each microfiche was deposited at the Census Room for public use.

In 1995, it came to my attention that the problem of reading very damaged documents had appeared to have been solved by the authorities in Spain. A computer software package which incorporated the use of a document scanner and imaging camera had given very positive results from fifteenth-century documents that appeared at first sight to have the same type of damage as shown on the Manchester returns. Examples of "before" and "after" images from the Spanish documents gave a boost to the chances of an easier and more effective system being employed on the returns at the PRO and I wrote to the Keeper with details of and examples from the Spanish work. Unfortunately the PRO decided that the Spanish system was rather slow and labour-intensive and was unlikely to be of major benefit to our work. This was a disappointing set-back, especially now that we had reached the stage in our transcription work where the returns were even more difficult to transcribe, the easier ones having been tackled first.

As most readers will know, the Chancery Lane operation of the PRO closed in December 1996 and all functions except the Census Room and other small areas were transferred to Kew. It was important to maintain the continuity of our work at Kew, especially now that we could just about see the light at the end of a six-year tunnel. Realising that the availability of natural daylight was a prerequisite to the success of our work, my first venture into the dim area of the Editorial Services Department (facing due north!) filled me with trepidation and dismay. In fact, Kew One, where the ESD was located, was deliberately designed to reduce the amount of natural light available in order to protect the original documents. However, a knight on a white charger (in the form of Mario Aleppo, Head of the Conservation Department) came to the rescue.

In spite of the need for external light, Mario's first task was to show me a location in his department with no windows!! Ironically called the Disaster Room and normally used whenever documents were in need of special recovery treatment, this room eventually proved to be absolutely ideal for our work and we are most grateful to Mario Aleppo for his foresight. He demonstrated on the unfilmed returns the use of the latest in ultra-violet light aids - a hand-held scanner - and convinced me beyond doubt that reliance on natural daylight to read the unreal was a thing of the past. In addition to these facilities, we were allowed to use artificial light for those occasions when UV couldn't pierce the gloom of washed-out ink. Our experience over the past eighteen months working in the Disaster Room has confirmed that a combination of UV and artificial light sources have usually uncovered those parts of the returns that can be read. Because we are now approaching the end of our task, the returns that we are currently transcribing are the worst of all.

This was a deliberate policy so that we could build up an expertise on the more readable returns, ready to tackle the difficult ones. The current ones are producing retrieval rates of less than 50% due mainly to extremely weak ink being used by the enumerator, with the consequent loss of definition from water damage. Because of the poor condition of these returns we are making a single transcribed copy but are spending more time on a careful recheck of the retrieved data.

Our technique on the returns now being worked on consists of the following:

1. The returns are checked page by page using a strong artificial light source and a magnifying glass. Each column is checked vertically rather than checking each line horizontally. In this way, we are able to identify family names, relationships, ages and birthplaces with more accuracy.

2. If there are areas that have not been transcribed by the above method, then a UV hand-held light source is used to scan the appropriate area of the page. We have to don protective eye cover when using the UV light and the room is in complete darkness to ensure the maximum contrast conditions. A powerful magnifying glass is again used at this stage.

3. The results are checked by a second team member. Particular attention is given to missing surnames, especially where all other details of the household have been found.

4. A Xerox copy is made of the completed transcription sheets and is held by myself as security in case of loss of the top copy.

5. The top copy is delivered to Manchester for the next stage - computer inputting. The final results are published by the society on microfiche.

The results of our work to date are set out below. The team has now reduced to four volunteers and we hope to finish the last batch of returns before Christmas with publication scheduled for mid 1999.

PRO ref
HO 107/
District Unfilmed population  
Original Recaptured %
2220 Chorlton on Medlock 6,400 100 <2*
2221 Hume 20,800 12,400 60
2222 Pendleton and Pendlebury 16,900 7,000 41
2223 Pendleton and Pendlebury 34,700 20,800 60
2224 Salford Regent Road 11,100 9,700 87
2227 Deansgate 6,000 30 <1*
2230 Manchester St George's 11,600 10,400 90
2232 Harpurhey, Blackley & Prestwich 9,000 8,000 89
2233 Ashton-u-Lyne & Knott Lanes 19,300 16,300 84
2240 Oldham-below-Town 5,700 4,400 77
  Totals 141,500 89,130 63
  *these originals were very badly damaged

What about our future work? We have learnt from our experience of using both artificial and UV light sources that it has been possible to decipher some of the most difficult returns. We therefore intend to analyse our past results and select those enumeration districts with a low recapture rate - below 60% - to carry out a recheck using our current techniques. This will further increase the overall recapture rate, hopefully towards 70%.

This project has been long and arduous and I would like to pay tribute to those members of the Manchester and Lancashire Family History Society who have assisted me with the work. Some have found the going rather tough and have only managed to stay with us for a few weeks; others, like Kath Arkwright and Jeanne Bryan have lasted the pace almost from the start and special thanks are due to both of them. We shall never know the full benefits of our work over the past eight years but, judging from the kind comments already given to us by those researchers who have discovered their previously 'lost' ancestors in the unfilmed returns, then these benefits will increase over the coming years.

Ray Hulley, Project Co-ordinator
December 1998

TopTop of Page

Copyright © 2000-2022 Ray Hulley. All rights reserved.