Search tips

Until the 19th century spelling of surnames was variable, so it is essential to check possible variants when searching for a name. To help you do this, this site has two powerful search facilities: the wildcard *, representing any or no letter; and the wildcard ?, representing a single letter. Using these carefully can enable you to track down variant spellings without having to sift through many irrelevant entries. For instance, if you were searching for Birch, and knew that it was often spelt Burch, you might try B?rch. But if you wanted all manner of Birchall variants, B?rch*l* would be suitable.

When searching for a surname before, say, 1600, greater thought has to be given to the possibilities. Mediaeval spelling tends to have unexpected doubling of consonants, and final e: so the surname Beal might appear as Beall or Bealle. Because the letter 'i' was not usually dotted in mediaeval handwriting, there is a tendency to use y, a letter less liable to be mistaken as part of the letter preceding or following: so Thin will appear as Thyn, Thynn, or Thynne. Some letters had different sounds in mediaeval usage, so the surname Judge would normally be spelt Jugge, which we tend to read as a form of Jug. Several mediaeval letters have dropped out of the alphabet: eth, thorn, wynn and yogh. On the other hand, the distinction between I and J and between U and V is a modern one.

Of the obsolete mediaeval letters, yogh was the most common, and represented a guttural sound now lost, but still shown by our spelling of 'knight' and 'daughter'. When printed and/or indexed the yogh may be represented as 'gh' or 'g' or 'y' or 'z' or even '3'. The form 'z' is most common in Scottish usage, and is found in surnames like Menzies, where the old yogh sound may still be heard.

Before 1200 surnames, such as there were, are often Latinized. Wood, for instance, will be found as de Bosco; Smith as Faber. Surnames that are derived from a placename commonly have 'de' in front, or 'del' if the place had a 'the' in front of it. Surnames derived from a nickname or personal description will often have 'le' such as 'le Fletcher'.

If searching for a modern name with De, Da, Du, Del, La, Le, St, Van, Vander, Ap, &c. it is worthwhile experimenting, trying the elements apart and joined together: for instance, Ap Rees as well as Aprees. Our indexes make no distinction for case: de Courcy is indistinguishable from De Courcy.

Mc and Mac may appear as such, or as M': just as the records vary greatly in treating Scottish and Irish surnames, it is as well to experiment with possible ways of representing these names when searching the indexes. For instance, Macduff might appear as m'duff, macduff, mcduff, mc'duff, mackduff, &c.

Lost Generations

Heritable surnames started to form in England in the 11th century, and were the norm in most areas by 1400. Patronymic surnames still changed generation by generation in Wales and in the far northwest of England (Cumberland, Westmorland and northernmost Lancashire) through even to as late as the 18th century, but the form of most surnames was set by 1300 in the rest of the country. Nevertheless, many surnames seem to appear in the record much later than 1300. Some, of course, belong to families that migrated to England later than then; but where are the others?

Some of the commonest English patronymic names, such as Johnson, are rarely found before the 16th century, as such. Most of the texts being in Latin, the earlier we go the more often Johnson is represented as 'filius Johannis', i.e., John's son, and as such one is always unsure whether the heritable surname Johnson is intended. It is impossible to tell from the record whether Willelmus filius Johannis de Newton was William son of John de Newton, or William Johnson of Newton. In fact the distinction is more in our minds than in actuality: to us the question is which was the name that passed to his descendants - Johnson or Newton - a question which was then still far in the future.

Similarly, occupational names will appear in a Latin form - Cocus or Coquinarius for Cook, or in Norman French le Keu. Where a name is shown in its English form, the normal mediaeval English spelling may be rather different from the modern: so Judge appears as Jugge.

If the surname of interest derives from a placename, you may expect as you go earlier for it to converge with the mediaeval form of the placename, which may not be instantly recognizable from the modern. This is most vividly shown by the numerous placenames along the river Thames, all of which started with the element 'hythe', but became Chelsea, Lambeth, Putney, Maidenhead. As the placenames mutated, they commonly gave rise to variant surnames that look radically different, such as Goldstraw and Goostrey.


  1. If the surname is translatable, such as a patronymic, occupation, or common noun or adjective (such as Spring or Black) bear in mind what the Latin (and Norman French) equivalents are.
  2. If the surname is an occupation, common noun or adjective, check what the normal orthography was in the Middle Ages. This also applies to parts of a name; -head will appear earlier as -heved or -heued.
  3. If the surname is from a placename, find out from the appropriate Place Name Society volume or Ekwall what the earliest forms were.

Apprenticeship Secrets

Genealogists are familiar with 18th-century apprenticeship records, but more can be wrung from them than appears at first sight.

Any man who appears practising a trade will have learnt it from his father or from an established master. Practising a trade in a town or city was, in theory, the preserve of the freemen of the borough or city.

  1. So, the first step if the ancestor in question was trading in a town or city is to check the freedom rolls. Freedom was gained (apart from a few cases of grant or purchase) either by being son of a freeman, or having served an apprenticeship to one. The freedom roll will state the name of the father or master, and as freedom was usually taken up about the age of 21, will also give a rough indication of birthdate. Apprenticeship was commonly for seven years, say from 14 to 21, but occasionally started earlier or later.
  2. Charitable or poor-law apprenticeship bonds, if surviving, will be found in the appropriate parish registers. Some towns and cities kept registers of such apprentices, and these registers usually stated date, master's full name, trade, period of service, apprentice's name, his father's name, and abode.
  3. All non-charitable apprenticeship indentures became subject to a stamp duty of 6d in the £ (or 12d if registered late, or if the premium was £50 or more). The registers of these indentures survive from 1711 to 1811. They are now in the National Archives in the class IR 1, and are largely covered at These are particularly attractive to genealogists because the early registers normally give the apprentice's father's name and occupation: but by the 1750s entering those details became more fitful, and by the late 18th-century it was the norm to omit them. There are two parallel series of registers: one for duty paid directly in London, and one for returns from the country collectors. The former series is sometimes called the City Returns, but that may mislead, in that it included not merely London apprentices, but many from all over the country for whom duty happened to be paid on London. A similar trap is the assumption that the collector for a particular county only dealt with apprentices from within that county: a notorious example is Northamptonshire, which also includes most of the returns from Birmingham. An entrepreneurial spirit seems to have reigned within the collectors, who were happy to bring in fees from far afield.
  4. The IR 1 registers were essentially a record of cash received, and are set out on facing pages (two matching scans) with the details of master and apprentice on the left, and the date of the apprenticeship indenture, term of years, premium paid, and tax paid, on the right. There are thus two dates recorded, three in the case of the country returns. On the far left is the date that the money was received; on the right-hand page the date of the indenture (recorded as a check that the tax was paid promptly); and in the case of the country returns the date of the collector's warrant, i.e., the date that the money was paid in in London. An important point about this is that, even assuming prompt payment of the stamp duty, the date of the receipt of money in London from the country collectors would often be a year or two later than the date of the apprenticeship indenture.
Having said all this, what extra can be gained, even allowing for the happy event that the ancestor's apprenticeship record, better still with the name and abode of his father, has been identified?

  1. It is noticeable, looking through the IR 1 returns, that the apprentice's surname is often the same as the master's. In fact, the apprentice's master is often an uncle or cousin. So, in all cases, having found your ancestor's apprenticeship, it is as well to look closely at the master, and discover what you can about him and his family, with an expectation that there may be a link. And a good clue as to whether the master is a relative is if the amount paid for premium for the apprenticeship is rather lower than the going rate for the trade in question.
  2. Another phenomenon that occurs in the registers, most obvious in the country returns, is a grouping together of surnames. Where already know that a set of apprentices were brothers, rather than them coming into the registers each a couple of years apart, as they each reached a suitable age, they may be apprenticed within a month or two of one another. The reason for this is the very common practice of wealthy and/or childless uncles and aunts to make provision in their wills of sums of money to be paid in premiums for all their nephews to be put apprentice, as being a sensible way of setting them up in a trade for life. In consequence, shortly after probate has been granted, two or three or more of the nephews appear in the apprenticeship registers, within a few months. Particularly with the country returns, therefore, having found the apprenticeship of your ancestor, it is worthwhile looking very carefully at the names of the apprentices immediately preceding and following. And, of course, having detected such a set, there is then also the originating will to find.
And if no apprenticeship can be found for your ancestor, a presumption arises that he learnt his trade from his father: in which case John Johnson wheelwright, is likely to be son of a Mr Johnson wheelwright of the previous generation.

When exploring the apprenticeship register indexes, don't forget that you can go to individual decades on the site, for instance for the 1770s, and then find an individual collector's return, such as for Masters of Apprentices registered at Aylesbury in 1770, and search it for your surnames of interest.

Census Secrets

The decennial censuses in England and Wales, giving personal returns from 1841 onwards, should be one of the surest routes from tracing ancestry: it is generally acknowledged that coverage was close to universal, so it should be possible to trace an individual back through the 19th century, finding him or her ten years earlier, same name, same birthplace, just ten years younger, and back into his parental family. In theory.

In practice there are many pitfalls. Some arise from particular circumstances:

  1. In the case of a married woman, unless her maiden name is known, the transition back to a girl in her parents’ household is not so simple.
  2. In many industrial areas, particularly South Wales and the clothing districts of Lancashire and Yorkshire, certain surnames are so common, and the range of christian names used so limited, that full name, age and birthplace are often not unique. Hosts of coal miners called John Williams or Thomas Jones thronged the valleys, almost impossible to distinguish from one another.
  3. Then there are the people who were away from home at the time of the census - in the army or navy, sailors, fishermen, prisoners, navvies, paupers in workhouses, and so on.
  4. And although a single nuclear family can usually be traced back census by census, the very large slice of the population that did the meanest jobs, servants, casual agricultural labourers, shop assistants in lodgings, are rarely at the same place ten years before, and their identification depends more on judgments as to the rarity of their name. If Solomon Thundercloud is aged 52, born in Shelton, in the 1861 census, there is reasonable certainty that he is the same Solomon Thundercloud, aged 42, born in Shelton, in the 1851 returns. But a John Smith, aged 52, born in London, recorded in 1861, will have several contenders in 1851.
So, if tracing back through the census returns is not always so easy, what can be done to solve the problems that arise? And is there more to be gleaned from the census returns than the simple snapshot of a household at gaps of ten years?

Until recent years the key earlier census returns - 1851, 1861, 1871 - were largely unindexed, more having been done on 1851 than any other, partly because it was released to the public earlier, but also because it was the earliest return to record relatively precise birthplaces, a chance to get a hook back into the parish register sources for people who were already old in 1851.

With the rise of the internet, there was a goldrush to stake out uncharted territory: census indexing was done rapidly, often outsourced to non-native speakers, and inevitably the results are patchy. An independent survey of internet census indexes found up to 40% mistranscriptions. Equally, although many of the original returns are calligraphic in their beauty, others were compiled in execrable scrawls, and would have been hardly decipherable by the writers on the following day. The fact that in many cases occupation names and common placenames were misspelt by the enumerators, and that the same surname can be spelt in two or three ways in the record of a single household, indicates that some enumerators were cavalier about such niceties.

If, then, it is unsurprising that you are unable to find a particular individual in a particular return, what can be done about it? There are two initial steps. Firstly, before the rise of the internet large swathes of the returns had been indexed by local family history societies: these indexes have in theory been superseded, but in the event that there was such an index for your area of interest for the missing year, it is as well to search it out. The people that compiled the indexes knew their territory well, knew what surnames were likely, were working at the records conscientiously and diligently, and are much less likely to have made mistakes.

The second thing to do leads us to the heart of what can be wrung out of the census records. Whenever researching any person in the past, there are four key elements to consider: family, house, job and religion. Genealogists’ cardinal sin is to concentrate on the first, family, to the exclusion of the other facets of life - which explains why so many ‘genealogies’ are fragments of unrelated pieces of ancestry spatchcocked together like fragments of DNA.

House, location, is almost as important as family. When you find someone in a census return, you find them not as a disembodied person, but where they lived, worked, worshipped, and had their friends - some of whom would be their present or future relatives. There, in the census return, you have laid out before you the intimate details of a whole locality.

A few decades ago, when the London census returns were virtually unindexed, a major genealogy company had traced a family for a client back to the 1851 census. The client also wanted to trace the maternal line, but all was known was the wife’s maiden name, her father’s name and occupation, and that she was born in London. It was clearly important to trace her father’s family in the 1851 London census, but there was no hint of where he lived: he did not appear in the trade directories. The client was told that the only way would be to work through the whole of the London census, a very expensive undertaking. The client agreed, the search took place, and her father’s family was found - living next door to her. This brings home vividly how important it is to make a note of who was living in the neighbouring houses when you have found an ancestor in the census returns. It is a simple thing to do, just a matter of a little bit of diligence, and it often proves invaluable.

Secondly, if a family has been found at an address in, say, 1861, and at different addresses in 1851 and 1871, you should always trace the original (1861) address in the 1851 and 1871 returns: again, the rewards from this little bit of diligence are often very great, revealing other parts of the same family at the one address.

But to do all this requires a precise understanding of where any address was actually located. Places change, house names change, street names change, and streets are often re-numbered. So the most important part of a census return, after the record of the household of interest, is the cover sheet, the first page of the enumeration book, because that specifies exactly the area covered by the enumerator. It may be as vague, in the countryside, as ‘The Township of Newton’, or in a town give a whole list of street names and numbers, mentioning various key landmarks such as public houses or churches. With the help of a contemporary map it is then possible to locate precisely not only where the ancestor’s house was, but the boundary of the enumeration district. That area, so delineated, is the first to look to for workplace, chapel or church, school, graveyard - all of which may have records relevant to your search.

Although the registration districts and sub-districts, being those used by the registrars of births, marriages and deaths, changed little in the 19th century, the enumeration districts, particularly in towns, were redrawn for each census, so where there is difficulty in locating a precise address in the next or preceding census it cannot be assumed that the enumeration district number will be the same: but, again, the cover sheet of the enumeration district books indicate precise boundaries - essential so that no household was omitted, and none counted twice.

In Victorian times there was a great surge in the building of Anglican churches in the cities - belatedly, in the face of a population that had been rising rapidly since the 18th century. Rather than create new Anglican parishes, ecclesiastical districts were formed, and the census administration was given the task of allocating population statistics accordingly wherever new ecclesiastical districts had arisen. This had to be done down to street level, and in consequence each enumeration district book cover specifies ecclesiastical district, and if more than one, which streets fell in which district. The same information is usually given on the top of each sheet of the return. These new Victorian churches - many of which have since become redundant - have baptism, marriage and burial registers too late to be duplicated by Bishop’s Transcripts, and too late to be covered by most computerized indexes. When they were brand new, spacious edifices they attracted huge congregations, had their own parish magazines and organizations - knowing the ecclesiastical district in which an ancestor lived is a first key to exploring this resource.

Hearth Tax Secrets

The Hearth Tax of 1662 was enacted 13-14 Charles II. c. 10. The mechanism of the tax was thus that each owner or occupier of any house, edifice, lodging or chamber was obliged by 31 May 1662 to deliver an account in writing to the constable(s) (or headborough(s) or tythingman/men) of all hearths and stoves within the house &c; the constable(s) then to immediately enter the house (during the daytime) to verify the return. Where there was no occupier, a written notice having been affixed to the door by the constable(s), they were to enter within six days to make their own assessment. Thus the tax embraced occupied and unoccupied buildings. (sec. iii).

The constable(s) were then to deliver all the papers - the returns from the households, their amendments and their own surveys - together with a list of the names of all such persons refusing to furnish an account, at the next quarter sessions following 31 May 1662 to the justices of the peace (sec. iv).

The justices were next required to cause the returns to be enrolled by the clerk of the peace for the county (or riding or division), and the clerk additionally to make a transcript on parchment, to be delivered into the Exchequer within a month from the receipt of the returns from the constables (sec. v). Hence it is that there are few areas of England for which the full 1662 hearth tax returns do not survive, either in the National Archives or in the county record offices. The first payment was due 29 September 1662 (sec. i).

No person who was exempt from church or poor rates by reason of poverty was required to pay the hearth tax (sec. xvii).

In addition, the parish poor law mechanism was invoked to identify further households inadequate to pay the hearth tax (sec. xviii). If the churchwardens and overseers of the poor, together with the minister of the parish (or any two of them, the minister being one) certified in writing to the two nearest justices of the peace that the house was not worth more than 20s a year upon the full improved rent, and that neither the householder nor anyone using the property had, used or occupied any lands worth 20s a year, nor had any land, goods or chattels worth £10 or more, that house was to be exempt from the hearth tax, and the householder was not to be listed in the return made by the constable(s) (sec. xviii).

So, apart from this last provision, which was at first rarely invoked but became commonplace later, the hearth tax returns should, in theory, give a complete list of householders, including those unable to pay the tax, in the summer of 1662. This is a key moment for those genealogists who have been relying mainly on baptisms, marriages and deaths from the parish registers to reconstruct their family trees, because most parish registers were disrupted - or are even completely lacking - from about 1640 to 1660, the Civil War and Commonwealth periods.

This is all very straightforward, but many an interesting fact has been discovered by looking at the parish register entries and the hearth tax returns in tandem. For instance, let us suppose that the ancestor in question, say John Smith, married to a Mary nee Jones, had children baptized in the period 1660 to 1670 in Newbury parish church. Does he appear in the 1662 hearth tax? He will only be there if he was a householder; what is more interesting than finding him is not finding him. Since he was definitely living in the parish, he was most likely living with his parents or his wife's parents. Clearly any Smith or Jones listed in the hearth tax becomes a likely father/father-in-law. Equally, if he does appear, that fact makes it much more likely that he had moved into the parish or that his father was dead. In fact, for a whole range of scenarios a careful consideration of the hearth tax returns (which also survive for some later years through to about 1670) can be very rewarding.