The Token Books of St Saviour Southwark
an interim search site
William Ingram, The University of Michigan, Ann Arbor
Alan H. Nelson, The University of California, Berkeley

Token Books: an Introduction
William Ingram

With the dissolution in 1540 of the Augustinian Priory of St Mary Overie, on the south bank of the Thames across from London, the priory church was newly dedicated as St Saviour and leased in 1541 (and later sold outright) to two small Southwark parishes, St Mary Magdalen (a parochial chapel attached to the priory church) and St Margaret (on the High Street), both of which were dissolved.

The new parish included within its bounds the Manor of Paris Garden, the bishop of Winchester's Liberty ('Clink Liberty'), and the western portion of the Borough of Southwark. It was home, in the later 16th century, to brothels, to bull and bear baiting, and to such playhouses as the Rose, the Swan, and the Globe. John Harvard, who endowed Harvard College in Massachusetts, was baptized there; Lancelot Andrewes is buried there; so is William Shakespeare's brother Edmund.

Like other parishes, the new St Saviour was expected to enforce the requirements governing communion, as set forth in the Book of Common Prayer, namely that everyone age sixteen or older should partake of communion three times a year, specifically at Easter and also at two other times. Names of communicants, and of non-communicants, were to be noted, for presentation at episcopal visitations. Churchwardens had to be ready with such lists, though practice seems to have confined itself to the keeping of names for the Easter communion only.

No doubt every parish kept such lists of names; but the lists at St Saviour appear to be the only ones to have survived. They run from the 1570s until 1643, in 144 books, listing each year - with some gaps where books appear to be missing - the names of all heads of household in the parish, along with the number of communicants in each household. In some years the names of the other communicants are also entered.

The books - comprising in total some 130,000 entries - are called 'token books' because the practice at St Saviour (as at other parishes) was to sell tokens to every parishioner age sixteen or older, initially for 2d each, then 3d, finally 4d. The books were originally intended as a record of token sales each year, though their value for us is as lists of names and what those names tell us about family reconstitution.

For new users of this site, some background information may be necessary to make the best use of the information in this website.  Much important evidence about tokens and their use comes from the parish of St Botolph without Aldgate, so the discussion that follows deals with that information before turning to the St Savior books themselves. 

I. Holy Communion.
II. Complying with the Regulations.
III. St Botolph: the Day Books.
IV. St Botolph: Communion Tokens.
V. St Botolph: Tokens and Bills.
VI. The Parish of St Saviour.
VII. The Form and Content of a Token Book.

I. Holy Communion.

The statutes governing communion are quite clear and unambiguous. Successive issues of the Book of Common Prayer in the sixteenth and early seventeenth century repeated, as did successive Articles of Visitation, the requirement that everyone should partake of communion at Easter and at two other times in the year. The words are unvarying from one prayer book to the next: ‘euery Parishioner shall communicate at the least three times in the yere, of which Easter to be one.’

Among the matters to be inquired into at successive episcopal visitations, as recorded in the various Articles of Visitation, is the matter of non-compliance with this regulation. In Matthew Parker's articles for Winchester diocese, from about 1575, the matter is put succinctly: his 34th article provided for inquiry by the visiting commission ‘whether any parishioner is negligent or obstinate in commyng to his owne parishe Church or Chappell for diuine seruice, and their receauyng the holy Communion at the least thrise a yeare.’

The Ecclesiastical Canons of 1603 were even more specific. Canon 112 required that the ‘Minister, Church wardens, Questmen and Assistants of euery Parish Church and Chappel, shall yeerely within fortie dayes after Easter exhibite to the Bishop or his Chancellor, the names and surnames of all the Parishioners, aswell men as women, which being of the age of sixteene yeeres, receiued not the Communion at Easter before.’

Notice was to be taken, then, of absentees from Communion; those who missed the Easter communion were to be reported directly to the Bishop, those who missed the annual total of three were to be reported at the visitation. The assumption underlying these requirements was that the wardens in the various parishes would institute some method of accounting that would enable them to keep track of who communicated and who did not. But were there in fact such procedures, and were they used? The answer is surprisingly difficult.

II. Complying with the Regulations.

One method of accounting for absentees may have been provided by the regulations themselves. The standard procedure set forth in the Book of Common Prayer required that ‘So many as intend to be partakers of the holy Communion, shal signifie their names to the Curate ouer night, or els in the morning afore the beginning of Morning prayer, or immediately after.’ Presumably the curate was intended to keep a record of such names, not only as a guide for estimating the proper sufficiency of wine and bread for the service, but as a key to attendance.

Different curates no doubt took this responsibility differently. One instance of how seriously the precept was taken may be seen in a memorandum preserved by the parish clerk of St Botolph without Aldgate, a parish on the eastern edge of the City. The clerk noted on 29 March 1584 that the curate, Robert Hayes, ‘Ded give warninge to all the parrishioners’ that ‘all that ded meane to Receyve ye holly comvnion should from hence fforthe com & give hem warninge therof the nyght before ... or els he wold not suffer them to go thervnto’. Further, the curate ‘ded meane to keape a Just note whether that the people do comvnicate thryse everie yeare accordinge to the queenes Maties Lawes.’

We should probably take this note at face value, as a statement of the existing regulations. That the curate needed to make such a statement may reflect a doctrinal or procedural rigidity on his part, or it may reflect his dismay at a general laxity on the part of his parishioners. In any event, the memorandum suggests that the curate had yoked in his own mind the twin notions of advance booking for communion and keeping track of non-communicants. The one may have been the means to the other.

A further note in the clerk's day book records that five years later, on 11 December 1589, a court was held in the parish church by order of the Chancellor of London ‘to reforme suche parrishioners as had not Accordinge to her maties Lawes Rsd the comvnion ffor the Easter tyme past.’ Presumably compliance was still falling short. The clerk's day book provides no information about the individuals concerned; it may well be that such lists were the prerogative of the curate and wardens, a further argument perhaps that the curate's lists were the basis of such information.

III. St Botolph: the Day Books.

There are other useful examples to be found in the day books of the clerk of St Botolph without Aldgate. These books have survived for several of the years at the end of the sixteenth century, covering most of the years of the clerkship of Thomas Harrydaunce, citizen and ironmonger, who held the office from 1582 to 1601. The books contain items of interest to the clerk but extraneous to the official Registers; notes on sermons, for example, or churchings, or the reading out of banns. The clerk also noted in the day book whenever a communion was administered, recorded how many persons communicated, who was the officiating minister, and how much was expended on bread and claret or malmsey for the service. The cost of the bread and wine was borne by the parish at all regular communions throughout the year except at Eastertide. The cost of Eastertide communions was met by the lay impropriator, or farmer, an individual to whom the revenues of the rectory, tithes and advowson had been assigned. George Clark was the farmer until 1597; he was succeeded in that year by Thomas Dow. They are the ‘mr Clarke’ and ‘mr Dow’ who appear so frequently in the records.

At St Botolph, communion was held once a month, on the first Sunday of the month, unless it was Eastertide; and the clerk noted in his day book that these were ‘monthly’ communions — ‘the monthely comvnion for Marche’ for example, which would be ‘at the Chardgis of the parishe according to Auntient custom’ (7 March 1596). These monthly services were the only regular communion services out of Eastertide except for the ‘newyeres’ services held on the first of January when 1 January was not also a Saturday or Sunday. The clerk did not otherwise think of a new year as beginning in January; though there were three ‘newyeres’ communion services on 1 January 1583/4, the clerk observed that the monthly communion on Sunday 1 March 1583/4 was ‘ye Last comvnion for ye yeare 1583.’

Eastertide was another matter. Beginning each year with Passion Sunday, and extending variously to Ascension Sunday, Whitsun, or even Corpus Christi Thursday, the communions were labeled as ‘Easter tyme’ communions. The communion on Passion Sunday each year would be labeled ‘A comvnion beinge the first that was ministred for the Easter tyme.’ During the period defined by the parish as ‘the Easter tyme,’ communion would be administered every Sunday, and often during the week as well, particularly the week on each side of Easter, and would be ‘at the Chargis of the farmar according to Awntient Custom’ (11 April 1596). On the average, one might expect to find fifteen or even more occasions for communion in the nine or ten weeks of ‘Easter tyme,’ with Easter Sunday itself providing five or sometimes six services.

From the successive entries in the St Botolph day books one can easily calculate totals for a few representative years, both annual totals and Eastertide totals. But one must have some sense of what these totals do and do not represent. If each adult parishioner was obligated to take communion three times in the year, then an assumption of full and total compliance might tempt us to speculate that the total number of individual communions administered over the course of a year would be three times the number of communicants in the parish. We might also be led to expect that the communicants served at Eastertide will account for a third of the total, and so be equal to the number of communicants in the parish. Such optimal conditions do not exist, of course; the hypothetical figures associated with them will be diminished in real life by parishioners who communicated fewer than three times, augmented by those few who communicated more than three times, and skewed unpredictably by the untraceable errors in counting or recording by the clerk. Such figures as we actually do have, then, may be a poor basis for extrapolation.

The fact is that at St Botolph the number of communicants on Easter Sunday, even at five or six separate services, did not equal the total adult population of the parish; nor, apparently, was it assumed that the requirement of Easter communion would be met on Easter Sunday itself by all the parishioners. The margin provided by a looser definition of ‘Easter tyme’ must have allowed for communion by all without straining the fabric of the church. In 1584, there were 743 communicants at the five services on Easter Sunday; but there were ten other ‘Easter tyme’ services that year, and the clerk noted opposite the entry for Whitsun (7 June) that there were ‘1831 in all for ye Ester tyme.’ By Whitsun 1600, sixteen years later, the seasonal total had risen to 2,170 communicants.

Further, no demonstrable relationship is likely to exist between the Eastertide totals and the annual totals. Instead of being a third of the annual totals, the Eastertide totals will be found in fact to approach or even to exceed two-thirds of the annual totals, raising the possibility either that many St Botolph's parishioners took two or perhaps all three of their required communions during the Easter season, or — and more likely — that numerous parishioners took one communion at Eastertide and stopped with that. It is thus arguable whether the annual total or the Eastertide total will bear a more meaningful relationship to the communion population, and thus to the total adult population, of the parish in any year.

One might try experimenting with the figures. One might assume on the one hand that the annual figures represent multiple — ideally triple — communions by most of the communicants. One might assume on the other hand that the Eastertide totals represent single communions by most of the communicants. These two assumptions will bring different results. The one will suggest a parish of close to a thousand communicants, the other a parish closer to two thousand. As all persons over the age of sixteen years were required to take communion, and as the persons in this category will probably outnumber somewhat the population under sixteen years, we might from these figures postulate a total parish population of either less than two thousand or less than four thousand souls, depending on which hand we prefer.

One practice at St Botolph that may have wider significance is the broad definition of ‘Eastertide.’ If the records of other City parishes had survived with equal fullness, no doubt the practice of defining Easter as a season rather than as a single feast day would be found to be widespread. There is, in any event, no reason to think of the practice in St Botolph as aberrant, for in other matters (as already shown) the minister and wardens were careful to observe the letter of the law regarding communion.

But other tests cast doubt on the validity of the figure of a thousand or even two thousand communicants. There are conventional formulae used by demographers for calculating mean population in a parish, as a multiple of mean annual christenings or burials, and while such procedures yield rough results it is somewhat disturbing to find that in the case of St Botolph they suggest a total parish population for these years in the vicinity of five thousand souls. So our estimate of a thousand or even two thousand adults from the communion figures must jostle with another estimate of over 2,500 adults derived by these other means. Perhaps christenings and burials are a poor index to population at St Botolph. Alternatively, perhaps a sizeable portion of the parish communicated once a year only, and still more abstained from communion altogether, as recusants or merely from indifference.

IV. St Botolph: Communion Tokens.

One must read through several years' worth of day book entries from St Botolph before encountering any mention of communion tokens. It is unfortunate for us that the clerk, otherwise so meticulous in recording the minutiae of parish life, found the token system so commonplace as to be undeserving of notice — at least until 2 May 1596, when for some reason he felt constrained to note that 108 communicants had been served that day ‘amongest wch nomber mr William Parteridge with Edward Parteridge & Richard Lowdin his srvants did receyve the comvnion wthout tokens & Edward Parteridge did give vnto me ijd for his token wch mony I did delliver vnto mr Clarke the tewsday after.’

From this entry we learn several things. Of foremost importance, we learn that tokens were required for communion at St Botolph. We also learn that one could partake without a token if one paid cash; that the cost of a token in 1596 was 2d; and that the income from the sale of tokens appeared to belong to the farmer, in this case George Clark. There is no suggestion in the entry that the use of tokens is a new procedure; and it is hard to imagine that Mr Partridge was the first parishioner of his generation to come to communion without a token. The safest assumption is that the parish clerk, for reasons unknown, elected in 1596 to begin making occasional notes about the use or non-use of tokens by the parishioners.

Once the notations about tokens begin to appear in the day books in 1596, they become a regular feature of the communion entries, but in each year the notations appear only against entries for Eastertide communions. One might cautiously assume from this that the tokens were an Eastertide phenomenon, and indeed later evidence will bear this out. The best examples of ‘token’ as against ‘non-token’ communions are to be found at the end of the Easter season in certain years. During the period covered by the clerk's day books, the limits of ‘Eastertide’ at St Botolph were not as constant as one might desire. Passion Sunday regularly marked its onset, but from time to time there was uncertainty about its proper termination. The norm seems to have been Whitsunday, but in some years confusion prevailed — and the record of that confusion serves as useful evidence for us.

In 1595, for example, the clerk noted that the ‘comvnion on Whitson Sonday being the viijth day of June’ was ‘ministred for a monethly comvnion’ and therefore ‘at the Chardgis of the parish’; there were ‘234 parsons’ present to receive communion, ‘of the wch nomber 97 had tokens.’ Those who had come with tokens were no doubt expecting to fulfill their Easter communion obligation on that day. The clerk's entry contains no comment about disallowing these parishioners, so perhaps for them the communion would count as Eastertide.

In the following year the clerk himself was not certain when Eastertide ended. He noted that 271 people received communion on Whitsunday (30 May) 1596, ‘but whether the said comvnion was ministred at the chardgis of the ffarmer or at the chardgis of the parish I do not knowe.’ The clerk thought it might be an Eastertide communion because, as in the previous year, there were people present ‘that had tokens for the Easter tyme.’ One parishioner, ‘Sara Moody a widow dwelling as she said in the libertie of Eastsmithfield,’ had come specifically with her 2d to buy a token ‘for that shee Receyved for the Easter tyme.’ Widow Moody presumably still needed to take her required Easter communion, and was not about to be told that Whitsunday was too late. The clerk apparently concurred, for he took her 2d, ‘which mony I did deliver to mr clarke the said day after Evening prayer.’

The communion for Whitsunday (4 June) 1598 had been ‘given owt by the ... Curat to be ministred for a monethly Comvnion’ and so it ‘was Ministred at the Chardgis of the parish.’ But not all the communicants agreed; ‘there weare that ded Receyve at the said comvnion wch had tokens 247 persons besydes there weare some 40 persons or thereabowtes wch had no tokens so that i take it that there Receyved in all at the said comvnion 287’ persons. Clearly the majority of the communicants expected an Eastertide communion, and brought their tokens accordingly; others took the curate at his word and came without tokens. The clerk noted in the margin of this account that the communion ‘Afterwards was by mr thomas Dow the farmar as I think payd,’ which would suggest that the farmer himself took it to be an Eastertide communion.

Sometimes the clerk himself made the error. On 24 April 1597, the fourth Sunday after Easter, he noted of the service that it was ‘the Last Comvnion that was ministred ffor the Easter tyme.’ He noted the following Sunday, Rogation Sunday, as a monthly communion, but also recorded the sale of tokens to ‘thomas hudson and his wyfe — 4d and to ‘Robert Bond & Sara his wyfe’ who ‘receyved wthout tokens Wherfore I receyved of him for mr Dow 4d.’ A fortnight later the clerk had recovered from his error, perhaps after conference with the farmer, for he noted the communion for 15 May ‘being whitson sonday which Comvnion was at the apoyntment of mr Thomas Dow being the ffarmar ministred for the Easter tyme wherfore it was ministred at his chardgis.’

In one year, for reasons not clear, Eastertide began a week early. The communion on 2 March 1600 — the fourth Sunday in Lent, and the week before Passion Sunday — should properly have been a monthly communion, but is noted in the daybook as the first communion for the Easter time. There may have been some confusion among the parishioners; only nine of them came to the communion, ‘whereof vj of them had tokens & three of them had no tokens.’ A gloss gives further information: ‘Jhon goulding & his wyfe And owld mr woodcock had no tokens but said they would fet there tokens afterward.’ Presumably the communion would count as an Eastertide communion for the Gouldings and old Woodcock, and therefore they needed to have tokens.

These instances suggest that the possession of a token was a necessary adjunct to communion at Eastertide, but not otherwise, and that in the grey area of Whitsun some parishioners might be fulfilling their Easter obligation and others not. It further suggests that parishioners with tokens and parishioners without tokens might find themselves side by side at the communion table, as was certainly the case in all the Whitsun examples cited. From this it must follow that the token served not as a ticket of entry to the service, but rather as a means of certifying compliance with the Easter communion requirement.

Nor was this true only of communion held in the church. Parishioners too ill or feeble to come to church might have communion served in their homes. On Saturday 22 May 1596 (eight days before Whitsunday) the curate served two families in this fashion, with the clerk in attendance. For the first family, Williams Seres and his wife, it was to be the Easter communion, ministered at home ‘becavse his wyfe was sick’; and ‘being for the Easter tyme I Rsd of him for there Tokens iiijd wch mony I did deliver to mr Clarke.’ For the second family, Christopher Sherfe and his wife, ‘she being verie sick or bedred at home,’ it was not the requisite Easter communion and so there were no tokens and no payment.

From these instances one can begin to build a picture of Eastertide communion at St Botolph without Aldgate, but much of the picture remains blank. We do not know, for example, how the generality of parishioners bought their tokens; nor do we know the precise means by which tokens served as evidence of attendance. It is of course an easy matter to sell tokens to parishioners, if all that is required is an exchange of money for tokens; and it is an easy matter to gather them in again at the communion table, if the fact of their collection is all that matters. Unfortunately, it is not at all clear whether the tokens were individually identifiable, so that they might be checked later in the vestry against a list of names. Their utility as evidence of communion requires some such identifiability, for a token can be proof of compliance only if it can signify unambiguously the communion of a specific parishioner. This is a vexed matter, and we will be helped only a little by a curious experiment tried in St Botolph for the first time in 1597, perhaps as the inspiration of the new farmer, Thomas Dow.

V. St Botolph: Tokens and Bills.

On Passion Sunday (13 March) 1597 the clerk noted that ‘the Easter tyme’ had begun, and that 41 parishioners had received communion. But, contrary to his usual practice, he listed their names in the day book, with each name followed by a number:
        Robert Chaplin 1
        Ric Greenowgh 2
        Jeames foster 3
        Alice Turner 4

 — and so on through the forty-one names. But the last number in the list is 48, not 41; and there are no names against which the numbers 15, 18, 26, 27, or 39 appear. Two of the names in the list — ‘George young 32’ and ‘Agnis Hood 34’ — have been deleted. With these exceptions there are, as the clerk noted, ‘xlj persons’ listed as having received communion. The list suggests that 48 persons were scheduled to receive, but that seven failed to appear; those seven had, however, apparently been assigned numbers.

And indeed the missing numbers do show up in subsequent lists. ‘Alice hampton 39’ heads the list for Palm Sunday, followed by ‘ffillis harryson 50’ and 205 other names for a total of 207 communicants. In this list the numbers go as high as 374, and against the last eighteen numbers on the list there are no names; ‘these wthout names,’ noted the clerk, ‘had tokens but there nams be not to me knowne as yet.’

For Maundy Thursday another 35 names are listed, the highest number being 416. For ‘good-ffreyday’ another eleven names are listed, with the highest number now 440, and the clerk further noted the presence at communion of four persons for whom he had neither number nor name. On Easter Sunday ‘Agnis Evance 15’ finally made her appearance, along with ‘Anne harrison 49,’ at the head of a list of 92 names — not very many for an Easter Sunday — and which the clerk has headed as follows: ‘the names of the Comvnicants that brought in bills for there Receyving the said 27th day of March ano 1597 and the nomber besyd wch Rsd by tokens weare as followeth.’

The clerk's distinction between the ‘names’ of those with bills and the ‘number’ of those with tokens is borne out at the end of the Easter Sunday list, where he noted the ‘Some by bills & as heare named 92 and by tokens 561’; and further on, ‘Some Rsd the 27th Day of march 1597 weare in bills & Tokens 653 Comvnicants.’ The clerk has made some distinction here which is not entirely clear. There were apparently 561 token-bearing parishioners present at the communion whose names were not recorded, and 92 other parishioners who had no tokens but instead had ‘bills,’ whatever they were. But the system of assigning numbers against the names is continued in this list. Where the Good Friday list had ended at the number 440, this Easter list goes only to 472; and indeed, all but ten of the names on the list have numbers attached that are lower than 440, that is, numbers that were assigned earlier than Good Friday.

On the day after Easter, 85 persons took communion, nine who ‘did Receyve by bills’ and whose names were listed, and 76 unlisted who ‘Rsd by Tokens.’ At the top of the list are ‘Robert Tuttie 26’ and ‘Mrs Tuttie 27.’ On the Sunday after Easter ‘George young 32’ and ‘Agnis hood 34,’ the two who had been deleted from the Passion Sunday list, make their appearance along with 17 others ‘per bills,’ while 113 others who received ‘per tokens’ are unnamed. And so on through the Easter season, as the names and the numbers struggle to retain some semblance of order and sequence. Of the first fifty numbers assigned, all but number 18 have been accounted for in these remarks; the bearer of that number may have failed to receive communion. The highest assigned number is 472, listed on Easter Sunday, and of the presumed 472 numbered parishioners, 419 are listed at some point during the Easter season. In addition, 977 parishioners with tokens are tallied but not listed. There may have been some overlap; Toby Wood, who may have tried to gain admission by claiming that he had a token, was required to be listed on Easter Sunday among the ‘bills’ because he had ‘no token’ upon him.

There is no more about this curious practice of bills, and the procedure was not repeated in subsequent years. As the tokens were presumably paid for in advance, so the bills may have been a kind of promissory note, a commitment to purchase a token, on the strength of which the parishioner would be allowed to receive communion. The system may have been designed to alleviate the problem of parishioners arriving at the communion table with their pennies, though the clerk's entries in earlier years do not suggest that this practice was burdensome. If this was the rationale, then the result of such a system would be that a parishioner could buy a token either before communion or after communion but not at communion. If it was to be after, the need for taking names becomes evident. The logic of assigning numbers is less evident, unless slips of paper with the numbers on them were turned in at the communion table in lieu of tokens, to be checked later against the list. This would have been a reasonably efficient system, and it raises once again the possibility that the tokens themselves were similarly numbered.

Indeed, the clerk's concern for matching names with tokens verged on the exemplary. On 2 May 1596 he noted that at the communion service ‘there was a tall man in a cloke of frentche greene wchsaid that he was mr Jons the brewers servant that did delliver me also ijd for that he had not a token wchmony I did delliver vnto Jhon Ivatt for mr clarke at the said comvnion but the man Did go his way wthout giveing me his name.’

VI. The Parish of St Saviour.

In the late sixteenth and early seventeenth centuries the parish of St Saviour's in Southwark was different in several particulars from the parish of St Botolph without Aldgate. St Saviour's was a larger parish, both in area and in population; administratively it was in a different county (Surrey not Middlesex) and in a different diocese (Winchester not London), and, though partly inside, it lay largely outside the boundary and thus outside the jurisdiction of the City of London. Further, it held title to its own rectory and tithes (and therefore had a rector rather than a curate), and owed nothing to any farmer or impropriator. Culturally, St Saviour's is important because of the playhouses erected there — the Rose, the Swan, the Globe, the Hope — and the large theatrical community that lived there, including Phillip Henslowe, Edward Alleyn, and (some would like to claim) Shakespeare himself.

It's easier to understand the complexities of holy communion at St Saviour's if one understands that the parish consisted of three segments, stretched east to west along the south bank of the Thames. The eastern end was the most heavily settled, and was anciently called the Borough, or Boroughside; it enclosed the foot of London Bridge and extended along Borough High Street as it ran southward to Newington. The parish church itself — known as the old priory church of St Mary Overy before the dissolution — stood (and still stands) in the Borough at the foot of the bridge. Boroughside fell wholly within the bounds of the Borough of Southwark, and was therefore the only part of the parish that belonged to the City of London. The two remaining segments of the parish were, at the western end, the Manor of Paris Garden, and in the middle, the Liberty of the Bishop of Winchester, commonly called Clink Liberty. Clink Liberty and Paris Garden were collectively referred to as the Bankside in distinction to Boroughside. Bankside was not part of Southwark (a point often misunderstood by social or literary historians).

But despite the difference in scale, St Saviour's and St Botolph were conformable with one another, and with the other London parishes that were their neighbors, in the regulation of parochial life. At St Saviour, as at St Botolph, tokens were used for Eastertide communion, and Eastertide was defined broadly in both parishes. And, though nothing like a clerk's day book has survived at St Saviour's, a more extensive and various kind of information about communion is to be had from the series of account books, extending from the 1570s to the Commonwealth, in which the sales of Easter tokens to parish households were recorded. These books seem to be complete lists of the heads of households in the parish of St Saviour's from year to year, arranged topographically, with notations about the number of tokens to be purchased each year for each household. In addition, marginalia in several of the books seem to offer evidence about the numbers of tokens received at various communion services, their cost, and even the means by which they were distributed.

It is important to stress that the books ‘seem to be’ these things, for it is all too easy to allow the sweet reasonableness of certain propositions to carry them imperceptibly over into certitude. This is an especial hazard for those pondering administrative procedures, where modern notions of rational process may be quite at variance with the norms of sixteenth century practice. We might well run the risk of falling into such error if we were to plunge directly into a discussion of the books themselves, based only on our brief survey of St Botolph. It might be safer, and a better service to the truth, if we were to begin at one remove, and consider the token books as but one of several different kinds of account books kept by, or for, the wardens and vestry of St Saviour's parish. What the books were supposed to account for, and how they did so, will perhaps be better perceived if we look at the vestry itself and how it conducted its business.

We might begin by considering two important dates in the administrative life of the St Saviour's vestry. We learn from the vestry minute books that the election of wardens and other officers such as surveyors and auditors took place each year on 2 March (at St Botolph it was on 15 December). At this meeting the newly chosen wardens assumed their offices; the outgoing wardens closed their accounts and turned over to the new wardens their books, keys, and stocks of money. So, for example, on 2 March 1595 the vestry minutes record that outgoing warden Thomas Howse ‘browght and deliuered into the howse 4 tithe bookes, vizt ij of Humfry Emersonnes, one of wm Pratts, and one of his owne; the two tooken bookes of his yeare, and the schoole accompte, which ... were deliuered to mr Haies one of the auditors.’

As 2 March is not a feast day of any significance, one must assume that the date was chosen for its administrative convenience to the vestry. The auditors, who were the first to receive the books and other accounts on 2 March, made their audit report to the new wardens, and to the vestry as a whole, on the quindene of Easter each year. This is a movable date, in contrast to the fixed date of 2 March, and again we need to assume that it was chosen for administrative convenience. The quindene of (or fortnight after) Easter might fall as early as Sunday 5 April, or as late as Sunday 9 May, giving the auditors from 36 days up to 70 days to present their report.

Perhaps a reasonable explanation is that the auditors needed a certain minimum space of time — say a month — to complete their work, but that any fixed date a month or so after 2 March would run the periodic risk of conflicting with the Easter days. As the whole of the church calendar at that time of the year was dependent upon when the movable feast of Easter fell, it may have seemed simplest to the wardens to make audit day a movable day as well, even if this meant that in certain years it might not arrive until well into May. Probably the wait did not matter. Nothing critical in the life of the parish was contingent upon the arrival of audit day, and indeed the annual reports seem to have gone smoothly enough, to judge from the accounts of them in the vestry minutes.

The choice of 2 March for election day was also perhaps related to Eastertide. At its earliest, Easter might fall on 22 March; Passion Sunday, which fell a fortnight earlier and which marked the onset of the Easter season, might thus be as early as 8 March. In such a year — and 1573 was such a year — the second of March would have been the Monday before Passion Sunday, conveniently close to the beginning of Eastertide but not encroaching upon it. Officers elected on 2 March would thus be responsible for the whole of the Eastertide activities in the parish, even in such a year as 1573. Closing the books for the previous year on March 2nd may have ensured — or may have been intended to ensure — that, no matter how early Eastertide might begin, the old accounts would be closed before its onset and new ones begun. In years when Easter was late — as in 1576 — election day would fall even before the commencement of Lent. Preparations for Eastertide might proceed at leisure in such a year, while in a year like 1573 they might have (or at least ought to have) proceeded in haste.

We have already seen one example of a warden, Thomas Howse, being held responsible for the Eastertide of his incumbency; Howse in 1595 turned in to the vestry, among other reckonings, the two token books ‘of his yeare’. Other parish accounts besides the token books ran from one March 2nd to the next. For example, two annual bills for communion bread and wine have survived; they each contain a year's worth of entries, beginning in March and continuing through the following February but not into the following March (P92/SAV/174 and /175). The entry limits for these bills thus match the limits of the term of office of an individual warden responsible for such accounts.

One parish activity in advance of Eastertide involved the preparation of the annual communion lists, or token books. The parish clerk, whose responsibility it was to prepare the books each year, was not constrained in his duties by the annual change of wardens. His own office had a longer term, and presumably he might, within reason, organize his tasks as best suited him. In years when Easter fell early, he might (if he were diligent) begin preparing the token books before 2 March; then again, he might not. Four times a year, on the traditional quarter days, the clerk submitted his account to the wardens, and some of these accounts have survived. For example, the accounts for all four quarters of 1613 are extant, and the first quarter account (P92/SAV/611) contains the following entries:
        Item     For rulinge & writinge the tithe booke     vjs viijd
        Item     For writinge the borough side token booke     iijs iiijd
        Item     For writinge the Bankeside token booke     iiijs

This account is dated 25 March 1613, four days after Passion Sunday. Accounts for the three other quarters of the year contain no reference to tithe or token books. So from this document we learn that in 1613 the preparation of the token books took place between Christmas and Lady Day, and also that the tithe and token accounts were separately kept.

Fourteen years later the books had grown in size and the cost of preparing them had risen accordingly. The clerk's bill for the first quarter of 1627 (P92/SAV/685) contains the following items:
        Item   For writeing the 3 token bookes   xijs
        Item   the twoe tithe books & ruleng them   vjs viijd
        Item   the writeing the twoe tithe bookes   xiijs iiijd

Easter fell on 25 March in 1627, so the token books were presumably done some time before Lady Day. Another quarterly account (P92/SAV/622), bearing no year-date but from about the same time, includes the following:
        The 3 token books finding paper   xiijs iiijd
        The Tithe books for books & all   xxijs

This account might possibly be from 1625, a year in which Easter fell late — on 17 April. Perhaps in such a year, finding paper for the token books was all that the clerk felt he needed to do by Lady Day. If we had the account for the following quarter, we could be more certain of this.

It seems likely, from these items, that the tithe books were of a roughly similar bulk and extent to the token books, and may well have resembled them in format. Certainly the wardens felt the tithe books to be more valuable to the parish; on 2 March 1596, following the elections, the vestry resolved that the tithe books, after they had been audited each year, were to be brought into the vestry and were ‘to remayne in the vestry for euer.’ Nothing was said about the safe keeping of the token books; it is thus curious that, despite this precaution, nearly 150 of the token books have survived, and not a single tithe book.

The conclusions to be drawn from these entries are that the parish clerk was responsible for the making of the books, and that he tried to do the work during the first quarter of the year, unless Easter was particularly late, but in any event before Passion Sunday. The Bankside token book for 1589 (P92/SAV/243) provides some confirmation of these assumptions, and offers the most precise dating of any of the documents. Across its front is written the heading ‘The Token Bouke ffor the Bankesyde wrytten The xiijth Daye of march Ano 1588’, i.e., 1589; the writing of this book was thus completed — or perhaps only just begun — a mere four days before Passion Sunday. If tokens were needed for the Passion Sunday communion, and if the book was needed to distribute the tokens, then the book could hardly have been completed much later.

The records are not clear about the precise identity of the persons responsible for the distribution of the tokens. The received wisdom has been that ‘the Wardens or their representatives’ took the burden, and this is sufficiently vague to be irreproachable. What is clear, from annotations on several of the token books, is that tokens were distributed to certain wardens responsible for the main divisions of the parish — Boroughside and Bankside initially, with Bankside further divided into Clink Liberty and Paris Garden after 1613. The wardens in their turn distributed the tokens, usually in lots of a hundred, to the persons who actually made the rounds.

Of the parceling out of the tokens there is some evidence. The Boroughside token book for 1640 (P92/SAV/233) has the name of John Hayman, Warden, on its front cover; and on the book's sixtieth and last page Hayman's nephew — perhaps himself one of the roundsmen — has left the following note:
        John Manby deliuered out the tokens
        her in the vestery for my vnckel debety haman warden
        one the 21the of march 1639
(i.e., 1640)

The scene in the vestry during this distribution was probably a conventional one. March 21st 1640 was a Saturday, and the day before Passion Sunday; one might expect to find more lead-time given to the roundsmen, but perhaps they began their rounds no earlier than was absolutely necessary. The practice in 1640 was, in any event, not substantially different from the practice half a century earlier. From the Bankside token book for 1589 (P92/SAV/243):
        Rsd by mr Tuchener & mr John Payne the xvth march 88
        in tokens ffor the bankesyde — iij C tokens
        Rd more the xvjth march 1588 — ij C tokens

Tuchener and Payne had been chosen wardens on the 2 March preceding, and may well have been acting as their own roundsmen for token distribution, the Bankside still being a manageable territory in the 1580s. But they began no earlier; 15 March 1589 — like 21 March 1640 above — was a Saturday, the day before Passion Sunday. Warden Hayman in 1640 may have been continuing a well-established tradition by starting on that day. The late start may well have been appropriate; for the rounds, once begun, continued for several weeks each year, as attested by several documents. A notation from 1633 may serve as illustration. On page 44 of the Clink token book for that year (P92/SAV/279) one such roundsman has written the following:
        Rd of mr wesson the 19 of Aprell 3 hundred of tokenes
        Payd the 20 of Aprell 2 hundred     2 - 10
        Payd the 21 of Aprell on hundred of tokenes     1 - 5 - 0

‘Wesson’ was Thomas Weston, warden that year. Three hundred tokens received from him, carried round and distributed, and the money (by this time increased to 3d per token) collected for them and turned in, all in two days. This is speedy work; but there is little reason for presuming that this roundsman's pace was unusual. To begin with, he had good days for starting; between 19 April (Good Friday) and 21 April (Easter Sunday) anyone should have done as well. But in the following week, 28 April to 2 May, it took him nearly twice as long to sell the same number. Thereafter his rate dwindled, until it rose again toward Whitsun (9 June). Nonetheless, he did his work; he diligently recorded the receipt of additional tokens, his own and probably Weston's as well, on subsequent dates; 300 on 28 April, paid for in equal installments on 29 April, 1 May and 2 May; ‘3 hundred moore’ on 2 May, and another 300 on 19 May, for a total of 1,200 tokens. We learn from these annotations not only that some roundsmen were diligent and speedy but also that they received their tokens on credit, returning money for them only as they sold them.

And so with the warden as well, one degree further up the scale. The warden received his allocation of tokens early on, and paid in his money only after he had collected it from his roundsmen. So in 1596 warden John Wrench received 2,200 tokens for parceling out; and on an undated Bankside token book that is probably from 1596 (P92/SAV/245) the following notation appears:
        Rs the 4th of Jvne 1596 of John Wrenche Chvrche
        warden for 2200 of tokens  — 18li - 06 - 8

The fourth of June was the Friday after Whitsunday in 1596, an appropriate time for a warden to be closing out his token account. And the notation, made perhaps by the clerk, would be of use to the auditors when Wrench turned over his token book to them nine months later on 2 March 1597.

That the responsibility for the tokens did not always fall upon the wardens is suggested by an annotation in the Boroughside token book for 1619 (P92/SAV/208), where it is noted that the tokens for Paris Garden were delivered to Mr Overman, a warden; those for the Clink to Mr Bucket, a vestryman; and those for the Borough to Mr Harvard, another vestryman. Overman may have been charged with distribution of the tokens for the entire parish — a task for which a warden would be a proper choice — and may have chosen the others as his assistants. In other years, names of more than one warden appear on some books, so there may have been no fixed practice.

Someone, perhaps the parish clerk, was charged with the ultimate custody of the tokens, from the time of their initial distribution to the wardens, and thence to the roundsmen and the parish, until the tokens were returned at the communion table and the money received for them brought back in to balance the account. No doubt the same person was responsible for maintaining the stock of tokens, replacing those that were lost or mutilated, and increasing the stock in response to a steadily growing parish population. But so far I have found no information to suggest how the parish was supplied with tokens. There are no extant records of the purchase of tokens from a lead founder or plumber or pewterer, or indeed any records involving lead for anything other than roofing the church. (The same is true at St Botolph.) Someone must have made the tokens, and perhaps even made them anew each year; in the Bankside token book for 1589 (P92/SAV/243) someone, perhaps the clerk, has reckoned the ‘Some of all the tokens mayd’ at 1,790. But the records in which such accounts were kept seem not to have survived.

A vestry document dating probably from 1613 (P92/SAV/794) contains a list of obligations under the general heading of ‘Busines that the 6 Churchwardens are annuallye or yearelye bounde to doe.’ The second and third items in the list are of particular relevance:
        2   In castinge tokens & takinge the number of Communicants
              4 or 5 daies yearelye are spent
        3   In gatheringe the tithes & deliueringe of tokens to 1442 housholds
              not lesse then 40 daies

The phrasing of item 2 is frustratingly vague. One wants to know what ‘castinge tokens’ signifies; is it a hurried locution for ‘casting up,’ or tallying? Or does it mean what it seems to say literally, the making of tokens? And what is ‘takinge the number of Communicants’? Does it mean simply reckoning how many parishioners will take communion? Or does it have to do with numbering the communicants, after the fashion of the numbered lists at St Botolph? And why are these two activities linked together in one item? Is it possible that the making of the tokens was in some way dependent upon the numbering of the parishioners? At present, certainty eludes us on these heads; but it is important to note the possibilities raised by the evidence.

Though the token books were intended to account primarily for the distribution of tokens, some few of them contain, in addition, detailed notes about the receipt of tokens at the communion table. In some half-dozen of the books this information is extensive enough to make it as useful as the corresponding kind of notations in the day books of St Botolph without Aldgate. If one tabulates the information in these notes one finds some surprising differences between the two parishes, however; where Whitsunday was at the ragged edge of Eastertide at St Botolph, it was the most popular of communion days at St Saviour's, with Whitsun communicants regularly outnumbering the communicants during the whole of Easter week. Further, the notes record that tokens were received at the communion table well beyond even Whitsun; in July, in August, one year even in November. These matters deserve closer attention, but will have to await another time.

VII. The Form and Content of a Token Book.

The books themselves were called token books by their users; they are nowhere referred to in the surviving records as communion books or Easter books, though nineteenth century scholars seem to have preferred to use such terms for them. The name suggests, though not in any conclusive way, that the main purpose of the books was to keep track of tokens rather than of communicants. The books themselves were made up of several sheets of paper measuring roughly 16 inches by 12 inches, gathered into a single quire and folded once lengthwise, to produce a book some six inches broad by sixteen inches high. A half dozen sheets would have sufficed for making a single book for the entire parish in the 1570s; by the 1640s three separate books were needed, one each for Boroughside, Clink, and Paris Garden, and twelve to sixteen sheets — each sheet being the basis for four pages — would be needed for each book.


Each token book was heavily used during its one year of service; it was carried repeatedly on the rounds, written in, edited, altered, and finally audited. The books as we have them represent the end product of all this use, and our first task is to make sense of the various kinds of notes and markings they contain. I have tried here to replicate the probable development of a typical book, using a section of an actual book as my model. This sort of conjectural reconstruction is of course risky, for all the reasons I've mentioned above, but the alternative is stasis; if one is going to be analytical one has to start somewhere. I have used as my example a segment of five names from the first page of the Boroughside token book for 1615 (P92/SAV/204). The image at the left shows how the segment looks in its present form.


But it did not come to look that way all at once. It began with the necessary sheets of paper assembled into a book, at which point the parish clerk copied names into it in a fair hand, working from his rough copy; in almost all instances the rough copy was the annotated and revised book from the year previous, although in a few instances separate rough books appear to have been made. The long and narrow format of the book made a single column of names and numbers most reasonable. The names are of the heads of households; the roman numerals are the wardens' estimates of the number of communicants in the households, presumably for the guidance of the roundsmen. The first stage of the token book, the completed fair list, would probably have looked like the reconstruction shown in the image at the left. Note the wide spacing between names; I take this as a sign that annotation and correction were assumed to be inevitable.

Needless to say, no document resembling this image exists; it is my reconstruction of the unmarked original by removing (via PhotoShop) all the accretions from the image above.


Once the new book had been prepared, it would have been given to the boroughside roundsmen to use as they distributed tokens and collected their pennies for the Easter season of 1615. For each token actually sold, a slash would be marked on the line joining the name to the estimated number. Usually the slashes can be seen to be in a different ink from the names; but sometimes the inks are identical. Did some parishioners buy their tokens from the parish clerk while the books were still in his possession? Did he mark some of the slashes himself? Alternatively, did he furnish the ink for the roundsmen on some occasions? The number of slashes would, ideally, agree with the roman numeral to the right, validating the wardens' estimates about the number of communicants in the household. Inevitably, of course, such was not always the case; it was true in three out of five instances with the names in our segment. With the sales made, and the slashes marked in, the segment ought to have looked like the image at the left.


Occasionally the roundsmen would note an inconsistency in the book, perhaps where a householder had moved away, or had died, or where someone else had moved in to join or replace another tenant. In such cases the roundsmen made the necessary deletions and changes in the book (sometimes introducing errors of their own). In the present case, two such changes have been made: ‘widowe Belcher’ has been deleted and the names ‘Richard dutton’ and ‘John blaxston’ written in, each in a different hand; and a name that appears to be ‘william wyntoun’ has been entered and then deleted. The result is something looking like the image at the left.


And toward the end of the year, when the final casting-up of the token accounts was made, in preparation for audit, someone — perhaps the roundsmen, perhaps the wardens or sidesmen — would have gone through the book once more to annotate any changes in the composition of the households listed, or to enter the names of new arrivals in the parish. These annotations would usually take the form of an additional number, most commonly arabic, written out to the far right of the page. This number represented the best estimate by the wardens of the number of communicants in each household for the Eastertide next to come — in this case, for 1616 — and would thus become the basis, or rough copy, for the new book to be made for the following year. After such annotations, the segment from 1615 might look like the image at the left.


Nothing remained, then, but for the parish clerk to make up his new book from this copy. When is this likely to have happened? Unfortunately we don't know. There is only one unambiguous piece of evidence, the 1589 Bankside token book mentioned earlier (P92/SAV/243), which has its date of composition written on the front: ‘The Token Bouke ffor the Bankesyde wrytten The xiijth Daye of march Ano 1588’ (i.e., 1589). As Easter fell on 30 March in the year for which this book was made, the date written there is just four days before Passion Sunday; perhaps making up the book so late was normal practice, perhaps not; but it may explain what was happening at the juncture between 1615 and 1616. Like 1589, 1616 was one of those years when Easter fell early, on 31 March; Passion Sunday fell on 17 March, and the first rounds may have been walked only a few days earlier. Parishioners wishing to buy their tokens more than a day or two in advance of Passion Sunday may have become statistics for the new book before the new book was ready for them. With such eager communicants accosting him in the church in the early days of March, the parish clerk may have had no option but to allow the old book to serve as an interim record of token sales for the following year. As a result, some early token sales for 1616 (in this case, Blaxston and Potter) show up in the old 1615 book as a second set of slashes at the far right. This is the final set of annotations inserted in the old 1615 book, and the resultant text resembles the document as it is found today.


And when the fair copy of the following year's book (1616, P92/SAV/205, shown at the left) was finally made, the corresponding entries for Blaxston and Potter show the same number of slashes as written in the old 1615 book, though Blaxton's name is now struck through. These slashes were presumably written in directly by the clerk, reflecting the purchases recorded originally in the 1615 book (see the image above). Note also that what happened in the previous year's book seems to have happened again with this year's book, with strokes appearing at the far right of the entries.

This much will perhaps suffice as a fairly risky general statement of the way the books functioned as records. But a lot of my assertions in the preceding section need to be substantiated, because they remain only hypotheses until validated. A number of them would seem to be self-evident, but that does not exempt them from the need for proof. The first assertion, that the names are of the heads of households, is conveniently demonstrated in the token books for the years 1620 to 1622. During those three years an effort was made to name and number every communicant in the parish, not just the heads of households. The project was abandoned in 1623, but for these three years the books are a rich if chaotic treasure-trove of information about parish families. As an example, Ralph Babington, one of the names in the previous example, whose household numbered seven in 1615 and 1616, was listed in the Boroughside books for 1620, 1621 and 1622 in the fashion shown in the next three images.

Entry for Ralph Babington in 1620 [P92/SAV/209]

Entry for Ralph Babington in 1621 [P92/SAV/210]

Entry for Ralph Babington in 1622 [P92/SAV/211]

The number of communicants in Babington's household grew in those three years from five in 1620 to eight in 1621 and nine in 1622, with the wardens predicting a return to five in 1623. One can see from these entries, however, that the bulk of those numbers were servants. These same entries support the claim that the numbers do not represent the size of the whole household but only the number of communicants. Thomas Babington filius (‘Tho:fil:’) appears for the first time in the 1622 list because he has reached communicable age; but the parish register shows the Babingtons with other children under sixteen, and with names different from the names of the servants listed.

An examination of the books for successive years will also show the relation between the ‘rough copy’ numbers added at the far right of the column in any given year and the fair roman number entered in the following year's book. The composite for Richard Yearwood, assembled below, is for the years 1613 through 1617; it fluctuates through several numbers, but the continuity of the changes can readily be seen.  The predictive number at the far right in any year becomes the roman numeral for the following year.






To test this hypothesis further, I compared several dozen names from the four Boroughside token books for the years 1612 through 1615 (P92/SAV/201 through /204), using all the names listed on the East Side to Ship Alley, Swan Alley, Goat Yard and Montague House. In only 25 out of 246 cases where it was possible to compare year to year did the roman number of the second year fail to repeat the arabic number of the earlier year. Some of these failures are explicable, others not. What did become clear as a result of my test was that the estimated number of communicants in any year was not based upon the number of tokens actually delivered the previous year; indeed, estimates continued to be given, and to change, even when no tokens were recorded as having been bought over a period of years. Even in a year when the number of tokens bought was equal to the estimate, a change in the estimate might well appear for the following year.

The information that produced the changes in these estimates thus must have come from some source external to the token books themselves. Perhaps there were annual lists of inhabitants required to take communion; or lists of all the parishioners, with communicants distinguished from the others. But no such annual lists are known to have existed for St Saviour's, unless they were included in the tithe books, of which none have survived. But if the tithe books listed all the communicants annually, then that list must have been based in turn upon another, so far unknown, list or — much more likely — upon personal knowledge or investigation. Lists of inhabitants made for other purposes — contributors to the poor or other rates, people in divided tenements, recipients of poor relief — would all have been partial and would have excluded some of an age to communicate. The likeliest explanation is that the estimates were made during the final walking of the rounds for the year, when the books were closed out in preparation for audit.

There is ample evidence that the wardens or their deputies took frequently to the streets. Mention has already been made of the document prepared in 1613 (P92/SAV/794) in which the work of the wardens was said to include the casting of tokens and taking the number of communicants, a task estimated to take four or five days a year, and the gathering of tithes and delivering of tokens to 1442 households, which was estimated to take 40 days a year. An annotation in the Bankside book dated ‘1593’, but which records transactions in 1594, makes clear not only a present ‘going about’ but an earlier one as well:

The 5th Janarye 1594 [i.e., 1595, as the book was being closed out] we ffynd to Receve of this bouke ffor the banke syde vc xxiij parsones besydes xxxiij tenyments the wch had no tenants in them when we went a bowt fferst    (P92/SAV/244)

Another annotation on the cover of the Boroughside book for ‘1597 & 1598’ (P92/SAV/190) is for the receipt of money for tokens ‘when we were a brode.’ And at the back of the Boroughside book for 1621 (P92/SAV/210), sums are noted as having been collected ‘the first dayes gowing’, on ‘the 2 day’, and on the ‘Last day’.

Token books apart, the wardens had other business to take them abroad. Draft churchwardens' presentments for ‘the liberty’, undated but probably from about 1640 (P92/SAV/587), indicate that there was a fortnightly or monthly perambulation by these officials, at least with the annual visitation in mind. From January 1592/3 (P92/SAV/450) they were to view newcomers and inmates to see if they were liable to become a charge upon the parish. Other officers or officials went about as well. The vestry appointed a surveyor for the sick and a searcher for the dead in May 1583 (P92/SAV/450), and the churchwardens' accounts for the poor suggest strongly that the parish poor were visited in their own homes. All these visits might have provided opportunity for the kind of annual ‘census’ necessary for updating the token books.

The evidence also seems to suggest that the men who delivered the tokens walked their rounds several times during each Eastertide season, and that token sales, and the marking of the token books, were ongoing activities each spring. A number of the token books contain annotations about the periodic delivery of tokens to the roundsmen, or of the receipt from them of their collected money. The normal pattern seems to have been for tokens to be parceled out in multiples of a hundred, and for the roundsmen to return their collected pennies within a few days and to take on more tokens. The most fully detailed account of the process is scrawled on the back of the 1633 book for the Liberty of the Clink (P92/SAV/279), from which the first few lines have already been reproduced. The ‘Wesson’ and ‘Inke’ of the document were probably the wardens Richard Inche, leatherdresser, and Thomas Weston. As the vestry minutes for these years have not survived, it is difficult to determine wardencies. ‘Dagoat’ may be Lambert Daggett, shoemaker; he seems to have been a roundsman for the Clink. The annotator himself is unidentifiable, but may have been the senior Clink roundsman. One can see from his account that he stood in some medial position between Weston and Inche on the one hand, from whom he received allotments of tokens, and Daggett, to whom he delivered them. Daggett turned in his money to the annotator, and the annotator paid Weston. The process continued all spring, well past Easter itself, with the annotator delivering a hundred tokens to Daggett for distribution just three days before the feast of Corpus Christi on 20 June. The account follows, in full, and these dates may be of some help in following the course of the action: in 1633, Passion Sunday was 7 April, Easter 21 April; 2 June was Ascension Sunday, 9 June Whitsun, 16 June Trinity Sunday.

        Rs of mr wesson the 19 of Aprell 3 hundred of tokenes
        payd the 20 of Aprell 2 hundred 2 - 10
        payd the 21 of Aprell on hundred of tokenes 1 - 5 - 0
        Rd of mr inke the 28 of Aprell
        3 hvndred of tokenes
        payd 29 of Aprell an hundred 1li - 5 - 0
        payd 4 of maye an hundred 1 - 5 - 0
        payd 11 of maye an hundred 1 - 5 - 0
        Red the 11 of maye 3 hundred moore 25s for iC
        payd the xij maye
        Recd moor in pt of 3C xxv s
        of tokens this 19th of maj
        delivared dagoat 25 of maj 2C
        of Toknes  - 26 may recd  - For 2 - 10s - 0
        2C of tokens of dagoat
        Res vn pd by dagoat for 1 C  - C
        deliverd on holy thvrsday  - 1 C ma
        Recd on holye thvrsdaye  - For 1 C  - xxv s
        so Rest for 1 C on thvrsday  - vn paaid  -
        deliverd on holy thvrsday noone 72. tokis
        deliverd the first of Jvn 1633
        To dagoat in Tokenes  -  - 200
        moar ij f vn bt dagoat  -  - 200 of toks
        Recd for 1 C moar of thise
        so Rest  -  -  - 1 C  -  -
        Moar on whitson eve ij C
        Recd on whitsonday of dagoat 3 C
        To dagoat on sondaye 9 of Jvn ij C
        Recd in pt of 2 C Recd  -  - 1 C
        Rest  -  -  -  - 1 C
        moar on saterday 16 iij C tokenes
        To dagoat
        Recd  -  -  - ij C
        Rest  -  -  - ij C
        Recd for this ij C
        17th deliverd moar  -  - 1 C
        (P92/SAV/279 p 44)

That the roundsmen both delivered tokens and collected money while on their rounds is made unambiguously clear by a note written by Francis Grove, a warden, in the 1632 book for Paris Garden (P92/SAV/309). Grove noted that he had ‘Rec in tokens to deliver at the first — 600’. As Passion Sunday in 1632 was on 18 March, the first going on rounds must have been in the previous week. Grove noted further that he ‘delivered backe the 15th of march’, a Wednesday, unsold tokens to the number of ‘317’ to his father in law Thomas Wycherley, a pewterer and thus perhaps a maker of tokens; and further, that he ‘paid him iijli xs ixd for 283, wchI had mony for’.

In another account book (P92/SAV/602), the clerk — or someone wishing to set out the relevant information — wrote the following notation:

  vppon  OblacionsThe Receipts of the
Weddingschurche doth yearely
Clark wages
the communion Table
And Rents

This is a useful list. Weddings and burials are self-explanatory. Knells are the tolling of bells at funerals, for which a fee was payable. Oblations and offerings are terms so broad as to be unusable. Rents and tithes are documented in the churchwardens' accounts, though not in much detail. Clerks' wages were commonly assessed against a parish as a separate item; evidence for such practice has survived for many London churches, for example St Bartholomew's by the Exchange, where the common rate was 2d to 6d per head of household. The important distinction in this list is between the income from tokens and the income from the communion table; this distinction cautions us to be wary of annotations on the token books about money received at the communion table, lest one should mistakenly confuse such money with token money. There are occasional marginal notations in the token books about receipts ‘at the comynyon tabell in mony’ (P92/SAV/244), but the sums so noted are themselves proof in part against this kind of error, for they are rarely exact multiples of the 2d or 3d cost of a token. Sums ending in three farthings (1593) or an odd penny and halfpenny (1598, 1599) are the norm.

The rough accounts for 1628 of a Mr Marshall, probably a warden, have survived (P92/SAV/694). The accounts include a breakdown which illustrates some of the distinctions made in the list just mentioned. Here is the relevant portion (‘borrow’ is Boroughside):

        in the borrow book                       66 08 9
        in the banke side & clinke lib   046 17 4½
                                                            113   6   1½
        for Clarkes wages in the burow boke Rec   13 16 1½
        ffor Clarkes wages in the bankeside & clinke   05 15 6
                                                                                19   12   1½
        in the 3 token bookes for the whole yeare   60 00 00
                                                                      4800 . at . 3d

These entries show that collections for tithes and for clerks' wages were both entered in the tithe books. The corollary should be that only token accounts appear in the token books; yet the token books contain such a haphazard array of annotations, from tithe accounts to details of presentments, that one may fairly assume the lost tithe books to have been equally accommodating. Indeed, the 1628 token book for Paris Garden (P92/SAV/304) has a set of figures on its cover which appears to be a recap of Marshall's tithe sums above. Similarly, the missing tithe books may have contained some useful information about tokens.

It seems clear from the foregoing that token sales were a kind of levy upon the parish, in much the same way that tithes and clerks' wages and the poor rate were. Much care was expended upon identifying the persons subject to the levy; the token books bear witness to that if to nothing else. They also bear witness, in an indirect way, to what appears to be a relative lack of concern about the return of tokens. Token sales were carefully and individually tallied; token returns were noted in bulk, in aggregate, and no tally seems to have been kept in the books themselves about who returned tokens and who did not.

We know from other sources that the wardens were expected to keep track of those parishioners who did not meet the mandatory requirement for communion. There was surely some system at St Saviour's for doing so; but the token books do not seem to be the place where the records were kept. The occasional notation about non-attendance serves rather to heighten than to mitigate this deficiency; for example, in the Bankside book for 1606 (P92/SAV/255), it is noted of widow Gray that ‘she receved nott in 14 yeres’ and that Roger Maggott's wife ‘received nott in 3 yeres’. The implication seems to be that these flagrant violations have only just been noticed. Perhaps the comments, like other comments about misbehavior that are scattered through the books, were intended to be transferred to some more appropriate book against the day of presentment; but no such other book has survived.

Nevertheless, there are some few books included among the series of token books that seem to be records of persons who have not taken, or who have not returned, their tokens.

During the years 1620 to 1622, when the experiment of naming and numbering all the communicants was tried, there may also have been a system of bills or tickets instituted, similar to the system at St Botolph three decades earlier. In the Paris Garden book for 1621 (P92/SAV/299), Richard Stevens and his wife are numbered 380 and 381; and at the back of the book is a note that ‘Rychard scalens saieth hee had tycketts and did deliver them in at ester’ and that ‘the numbers’ were 380 and 381. Whether or not Scalens is an error for Stevens — and it probably is — the important clue in this note is the mention of numbered tickets. The note also suggests that someone had challenged Stevens or Scalens about attendance at Eastertide communion, and that he was ready with what must have been the proper reply; citing his ticket numbers as proof. If indeed such numbers constituted proof, then they must have done so in some other book, for the token book merely records that he was assigned those numbers, not that he made use of them.

The tokens were returned by each communicant, at the church, on the day the Eastertide communion was taken. The token books are silent about these individual returns. There is a note in the book for the Boroughside for 1636 (P92/SAV/228) of the total for the three divisions of the parish (5,379), of those ‘rec' backe’ (5,296), and of the difference, ‘rec' not 0083’. The total ‘in this booke’ for the Boroughside matches the total number of slashes, and so for Paris Garden as well (P92/SAV/315). This would suggest that the tokens in hand were counted against the number delivered, not against individuals. In the book for 1602 (P92/SAV/195) a list was made of ‘Tokens receyved back again’ between 20 March and 25 April 1602. The first two entries are for tokens ‘back at communion table’; the brevity of the other five entries suggests that they, too, were returned at the table.

Other evidence points in this same direction but cannot conveniently be checked in the same manner. There is a note on an earlier book for the Boroughside for 1593 or 1594 (P92/SAV/187), ‘the 5th of January 1594 we ffynde to receve of this bouke iij C Lvij parsonnes’, which may be a count either of communicants who have not taken tokens or who have not yet returned them. In 1595 (P92/SAV/191) it was noted ‘In this bouke the 25th July 1595 to receve & tacke tokens — 570 persons’.

The presence or absence of individual parishioners at communion could have been affirmed by any number of other people. The presence of parish officers and their helpers during divine service is clear. In March 1614/5 the vestry ordered that an officer should stand at the doors between church and chancel to supervise the entry or exclusion of parishioners until the beginning of the sermon (P92/SAV/450) and in 1639 (P92/SAV/520) it was stated that parishioners were to be seated by the wardens and six to eight ‘ancients’. Such men could take the return of tokens, particularly if no check was to be made upon who was returning them. But whether they did or not is now lost to us.