Hostname: page-component-6766d58669-7fx5l Total loading time: 0 Render date: 2026-05-18T20:37:56.626Z Has data issue: false hasContentIssue false
  • English
  • Français

When Did Augustine of Canterbury Die?

Published online by Cambridge University Press:  10 June 2016

RICHARD SHAW
RICHARD SHAW*
Affiliation:
Our Lady Seat of Wisdom, 18 Karol Wojtyla Square, PO Box 249, Barry's Bay, Ontario, CanadaKOJ 1BO; e-mail: rhshaw@seatofwisdom.org
Rights & Permissions [Opens in a new window]

Abstract

While Bede did not know the year of Augustine's death, he possessed papal letters which provide sufficient information to deduce it with some confidence. The early epistles from popes which Bede quoted or referred to in the ‘Historia ecclesiastica’ associated journeys by delegations sent by the early Church in Kent to Rome with the request for, and collection of, the pallium for the new bishop of Canterbury. In this light the likely purpose for the otherwise unexplained visit of Mellitus to Rome in 610 becomes clear: he had come to ask Pope Boniface IV for the pallium for Laurence, following the death of Augustine on 26 May 609.

Information

Type
Eusebius Prize Essay
Information
The Journal of Ecclesiastical History , Volume 67 , Issue 3 , July 2016 , pp. 473 - 491
Copyright
Copyright © Cambridge University Press 2016 

Despite the crucial role that the Gregorian mission plays in the Historia ecclesiastica, acting essentially as the keystone of the book's entire narrative, there is a surprising amount of basic detail about the early Church in Canterbury that Bede apparently could not discover. One frequently-noted absence from the HE, for instance, is the year of Augustine's death. Bede tells us that he died on 26 May, but he does not give the year.Footnote 1 Bede is quite open about his source: the Canterbury epitaph for Augustine. Indeed, the calendar date is included in the HE only as part of the text of that epitaph, which Bede quotes in full. But because the epitaph did not note the bishop's year of death, Bede did not know it either. Unfortunately, as with so much about early Christian Anglo-Saxon England, Bede's ignorance usually means our ignorance. Some historians, faced with this lacuna and wanting to go beyond the mere range of possible dates, 604 × 610,Footnote 2 have guessed or presumed what Augustine's year of death might have been, claiming that it fell in or around 604. These assumptions are based on the position of Augustine's epitaph in the HE's narrative. Bede included it at the end of HE 2.3, which begins with the description of Augustine's consecration of Mellitus and Justus and the foundation of the sees of London and Rochester in 604. This textual propinquity has then been taken as a sign of chronological connection, leading to the presumption that Augustine died the same year.Footnote 3 Such an association of material in a chapter is, however, a very poor basis for deductions concerning Augustine's actual date of death, especially as Bede himself did not know in which year it occurred. One cannot treat the mere relative placement of the event in Bede's work as evidence to support a specific year of Augustine's death.Footnote 4

None the less, the precise date, it turns out, is not lost to us. The journeys of the early mission fathers to and from Rome provide clues which can point towards an answer. Careful consideration of the papal letters that Bede possessed elucidates the timings of and the reasons for the trips of representatives of the early English Church to Rome. Understanding the full context of these visits permits a confident conclusion that Augustine died on 26 May 609. The advantage of such an approach is that the conclusion can be based on primary sources rather than Bede's own narrative. The HE is not a primary source for the period. Bede possessed too few genuinely early sources for us to treat the HE's account of the early Church in Kent as if it were a contemporary record.Footnote 5 But the papal letters that Bede used are the notable exceptions. They are the most reliable basis for reconstructing events in the early seventh century.

Papal letters and visits to Rome

As he informs us in the preface to the HE, Bede obtained the papal letters that he quoted, or simply drew from, thanks to the research in the Roman archives of Nothelm, priest of London and future archbishop of Canterbury.Footnote 6 Nothelm had brought the letters to Wearmouth-Jarrow with the encouragement of Albinus, the abbot of SS Peter and Paul, Canterbury, described by Bede as the ‘auctor ante omnes atque adiutor opusculi huius’.Footnote 7

The papal letters that Bede quoted and used, and those of Gregory the Great that have survived separately, are, of course, primary sources for events at the end of the sixth and early seventh centuries. They are therefore the best place to begin and the most reliable evidence for any attempt to reconstruct events in that period. One area about which the papal letters are particularly informative is the journeys to Rome by the early mission fathers. The letters enable us both to derive some details beyond those that Bede gives about the delegations sent to Rome, and to gain a good idea of the purposes of these visits and therefore about the succession of bishops of Canterbury. For these patterns and common themes to emerge more distinctly, it is necessary to collect together those letters or references to letters which reveal journeys from the early Church in Kent to Rome and to set them alongside each other.Footnote 8

There were five visits to Rome by delegations from the early English Church for which there is evidence prior to the episcopate of Deusdedit.

Visit i, 600–1

Bede inserted into his narrative several letters from Pope Gregory the Great sent in the summer of 601. These include letters commendatory for the party to take back with them to smooth their journey, such as that to Bishop Vergilius of Arles, quoted in HE 1.28,Footnote 9 as well as missives for Augustine, King Æthelberht and others on the party's return to Kent. Examples of the latter are integrated into the text of the HE at 1.29, 1.31 and 1.32.Footnote 10 In addition, in HE 1.30, Bede inserted a papal letter to Mellitus.Footnote 11 This was not written or sent at the same time as the others, but was composed about a month later and was sent to Gaul to reach Mellitus as he travelled to England. Bede also shows signs that he possessed other Gregorian letters from this period: for instance that sent to Æthelberht's wife Queen Bertha,Footnote 12 and perhaps one or more sent to other Frankish correspondents.Footnote 13 Other letters that Gregory wrote during the same period for the party to deliver to addressees en route to smooth their journey, of course, survive; if there had been any doubt these would confirm the trip to Rome and the group's return in 601.Footnote 14

Although all of Gregory's letters are from the summer of 601, the visit is better dated to 600–1 because the logistics of journeys to Rome at the time meant that the party probably initially set out from Kent in the summer of 600. Communications between Rome and Gaul in this period took place by sea and avoided the winter months as is evident from an examination of the entire corpus of the letters of Gregory the Great. Almost all of the pope's missives to Gaul (and those that went via Gaul, for instance to Spain or England) date from the late spring and summer, that is, from May to September, with a few from April marking the earliest point at which letters went sent.Footnote 15 Gregory only wrote two in October,Footnote 16 which therefore seems to have marked the end of the sailing season for practical purposes – unless there were exceptional circumstances, such as those surrounding the letters written in November 602 where Gregory appears to have been facilitating urgent negotiations between the Franks and Byzantines.Footnote 17 Even so, the winter months are completely avoided in correspondence with Gaul in the Registrum.

Crossing from the Italian coast to Marseilles could be dangerous at the best of times. The only logistical detail that Bede gives from an early journey between Rome and England confirms this: in HE 2.20 he states that Romanus of Rochester drowned in the Italian Sea while on an embassy to Pope Honorius.Footnote 18 King Reccared of the Visigoths claimed that his envoys to Gregory the Great had almost died on the journey and he apologised to the pope for their failure to arrive, so frightened had they been by the condition of the sea at Marseilles.Footnote 19

Still, in the early seventh century a sea voyage was better than passing through Lombard territory. Within a few decades, of course, the situation would change significantly. Following the dramatic emergence of Islam onto the Mediterranean scene, as well as extended periods of peace between the Empire and the Lombards, the overland route, making use of the Alpine passes, began to predominate among journeys between Rome and Gaul, though not to the total exclusion of the sea route when absolutely necessary.Footnote 20

In winter, however, as Gregory's letters plainly imply, conditions seem to have ruled out travel altogether. This is not to say that all maritime communication from Rome stopped: Gregory's letters show that boats could hop up or down the Italian coast, for instance to Luni, Naples or Sicily, though much more infrequently than in the spring and summer.Footnote 21 Nevertheless most of Gregory's communications in winter tended to be with people in locations directly reachable by land, either to recipients in Central Italy or to the Lombards, or with correspondents where travel by road could represent the first stage of the journey: for instance, with Constantinople and other parts of the east. In contrast, the impact of winter weather on sailing conditions on the Ligurian Sea, between Marseilles and northern Italy, meant that this at least was considered too dangerous during the winter, at least on papal business. Between Gaul and Rome, therefore, communications were carried out between the spring and autumn and mainly in the summer.

The same is almost certainly true for communications between England and Gaul. It can easily be forgotten how dangerous a sea the Channel is.Footnote 22 Bede himself mentioned the storms which had hampered both Julius Caesar's invasions,Footnote 23 as well as the tempest that had almost sunk the ship bringing Germanus of Auxerre and Lupus of Troyes.Footnote 24 Almost certainly, therefore, journeys between England and Gaul in around 600, like those between Gaul and Rome, would only have taken place between the spring and the autumn, and most Channel crossings would have occurred in the summer, to take advantage of the more predictable weather. Except in exceptional circumstances the winter months would probably have been avoided altogether.Footnote 25

Even in these conditions, of course, it might still be considered feasible for a party to leave Kent in say late spring 601, travel to Rome, and leave for England by the early autumn of the same year, though such rapid efficiency would be surprising without good cause.Footnote 26 In this case, however, such a scenario seems particularly unlikely, if not impossible, given the June date of the papal letters.Footnote 27 Thus, the logistics make it more than probable that the party set out from Kent in late spring/early summer 600, arrived in Rome in the late summer/early autumn and spent the winter in the city before returning with the various letters, as well as the pallium for Augustine, in the summer of 601.Footnote 28

Visit ii, 609–10

Bede mentions letters from a journey to Rome almost a decade after the first one, although he does not include them in the text of the HE. In HE 2.4, he says that Mellitus, bishop of London, returning from a trip to Rome, brought back letters from Pope Boniface iv to ‘Archbishop Laurence, the beloved of God, and to all the clergy, as well as a letter to King Æthelberht and to the English people.’Footnote 29

Even though the letters are not included in the HE, and do not survive separately, Bede provides enough information to date them and the visit broadly since he says that Mellitus attended a synod while in Rome which issued its decrees ‘in the eighth year of the Emperor Phocas, on 27 February and in the thirteenth indiction’.Footnote 30 This converts to an incarnational date of 610. Thus Mellitus left England in the summer of 609 expecting to spend the winter in Rome and well aware that a return in the summer of 610 was the earliest that could reasonably be expected. Then, a few months after attending the synod in February, Mellitus began his return journey – probably in late spring or early summer 610 – with the papal letters that Bede, thanks to Nothelm, would come to possess more than a century later.

Visit iii, 619–20

In HE 2.7 Bede notes that Pope Boniface v addressed ‘scripta exhortatoria’ (‘letters of exhortation’) to Mellitus and Justus. As with those from 610, Bede does not include these epistles in the text of the HE,Footnote 31 but the context makes it clear that Boniface sent these missives having heard about Mellitus’ succession. Bede can only have known this from the content of the letters. Thus, necessarily, a delegation had gone to Rome announcing the accession of Mellitus, and had brought back the letters that Bede mentions here.Footnote 32

Although the date of the letters and, therefore, the visit, is less obvious than that of the 600–1 or 609–10 delegations to Rome, it can be inferred without too much difficulty. In HE 2.7 Bede notes that Mellitus died on 24 April 624 and that he had been bishop of Canterbury for five years. Earlier in the same chapter, Bede had given Laurence's calendar date of death as 2 February. Thus, Mellitus succeeded Laurence in 619 and dispatched a delegation to Rome that spring/summer. Given that Boniface only became pope on 23 December 619, the letters cannot have been written until 620 — probably in late spring or early summer, as usual. Therefore, the papal letters that Bede mentions emanated from a trip to Rome in 619–20. In this instance, it is evident from Boniface's date of accession, that, just as in visit two (609–10), the round trip cannot have been completed in one single calendar year, but must have spanned two, which was obviously the usual practice.

The only slight doubt cast over this dating is a mild ambiguity in Bede's reference in this passage to Boniface v becoming pope in 619. As Charles Plummer noted, Bede's language could conceivably be taken to mean that the letters were sent that year,Footnote 33 but this is not the natural sense of the Latin,Footnote 34 and since Boniface became pope on 23 December 619, the letters cannot have been sent that year.

It might seem, at first, somewhat surprising that Bede was able to calculate the year of Boniface's accession. Bede had no ready and reliable source of information for the years in which popes of this period died or when others succeeded. Both the Liber pontificalis, which Bede used frequently in the HE, and papal epitaphs, of which Bede possessed a collection as part of a sylloge,Footnote 35 did usually give the Julian calendar dates for the deaths of the pontiffs. None the less, while some manuscript traditions of the LP are known to have inserted indictional years of death for one or two popes, most did not and those that did were often inaccurate.Footnote 36 Bede rarely had sufficient independent or supplementary information to make such calculations correctly himself. Indeed, when he tried – as he did for his hero Gregory the Great – the conclusion he came to was wrong: he erroneously deduced that Gregory died in 605, whereas the correct year was 604.Footnote 37

None the less, in this instance, Bede probably did have sufficient information to come to the correct conclusion about the year of Boniface's accession. The letters that Pope Boniface sent can be assumed to have had a dating formula,Footnote 38 which, if the conclusion above is correct, would have amounted to ad 620. Since the original letters from the English Church will have arrived before Boniface became pope, it is likely that he mentioned this point in his own letters, just as pope-elect John did in the letter to the Irish that Bede quotes in HE 2.19.Footnote 39 With this detail, together with the precise date of the letters and the knowledge of Boniface's day of accession – 23 December – given in the LP, Bede had more than enough information to calculate the year that Boniface became pope.Footnote 40 This, therefore, resolves the only remaining question concerning the date of this delegation to Rome, which must have occurred in 619–20.

Visit iv, 624–5

In HE 2.8, Bede included a letter from Pope Boniface v to Justus, following the latter's succession to Mellitus as bishop of Canterbury. In this the pope explained that he was sending Justus the pallium – the ecclesiastical vestment and mark of honour which it had been customary for bishops of Canterbury to be granted by the pope since Gregory the Great bestowed one on Augustine in 601. Once more, therefore, it is clear that the English Church had sent a delegation to Rome which notified the pope of the episcopal succession at Canterbury. Here the existence of a delegation from Kent is even more unmistakable, as Pope Boniface's epistle to Justus specifically refers to having received letters from King Eadbald.Footnote 41

The letters to Edwin in HE 2.10 and Æthelburh in HE 2.11, although addressed to recipients in a different kingdom, are likely to have been sent at the same time, given the content that they shared.Footnote 42 Unfortunately Bede or Nothelm omitted the dating clauses from all of the letters.Footnote 43 Nor, unlike in HE 2.7 (visit iii), is there any other direct, separate, dating reference in close textual proximity in the HE which was probably derived from the letters.Footnote 44 It is, however, possible to deduce a convincing date-range for the epistles from the details that Bede provides together with information from other sources. The letters were sent by Pope Boniface v. He died on 25 October 625, so the latest the letters can have been sent is summer 625. Bede says that Mellitus died on 24 April 624.Footnote 45 Since Boniface is writing to Justus as Mellitus' successor,Footnote 46 the likelihood must be that a delegation set off from England for Rome in the summer of 624, following Mellitus' death, and began the return journey in late spring or early summer 625 bearing the letters, and the pallium for Justus.Footnote 47

Visit v, 633–4

The last visit by representatives of the early English Church to Rome before the accession of Deusdedit is easier to situate chronologically because Bede retained the dating formula at the end of one of the two papal letters sent back with the party that he included in HE 2.17 and 2.18. The letter of Pope Honorius to Bishop Honorius of Canterbury which Bede quoted in HE 2.18 ends: ‘Given on 11 June in the twenty-fourth year of the reign of our most religious emperor Heraclius and the twenty-third year after his consulship, the twenty-third year of his son Constantine and the third year after his consulship; in the third year of the most illustrious Caesar his son Heraclius, in the seventh indiction.’Footnote 48 Bede even converted this complex chronological construction into incarnational dating for his readers, saying: ‘That was in the year of our Lord, 634.’Footnote 49 As with the letters connected to visit iv, it is not only the existence of the letters, but also their content which makes it clear that a delegation had been sent to Rome. In his epistles Pope Honorius specifically refers to having received requests from both Honorius and Edwin for the pallium to be conferred on Bishops Honorius and Paulinus.Footnote 50

Of course, Edwin's defeat at Hæthfelth on 12 October 633 meant that he never received Pope Honorius' letter and, indeed, was dead by the time that it was written.Footnote 51 The date of Edwin's death does, however, conclusively show that the party cannot have travelled and returned in 634. They must have set out in 633, probably in the summer as usual.Footnote 52

Bede, therefore, possessed, and sometimes even quoted from, papal letters relating to five visits to Rome by delegations from the early Church in Kent. Taken together as a group certain links and associations emerge between these trips and the epistles related to them. These connections permit an insight into the recurrent purpose which lay behind each of these journeys.

The third (619–20), fourth (624–5), and fifth (633–4) visits all occurred in the immediate aftermath of the death of the preceding bishop of Canterbury and the installation of his successor. The party going to Rome therefore would have notified the pope both of the former bishop's death and of the new bishop's succession. But this was not reason enough to make a trip to Rome. The papacy did not require or receive notification of every episcopal succession in the Western Church in the early seventh century. At this stage a delegation was no more likely to be sent to inform the pope of the election of a new bishop of Rochester than was one for the election of a bishop of Nantes in Gaul. So why did parties set out following episcopal succession at Canterbury? It was not because from 601 Canterbury was a metropolitan see. Such a status did not mean that Canterbury had to keep Rome informed about changes of bishop, any more than other metropolitan sees such as Tours felt the need to update the papacy on succession in their diocese. Successors were chosen and consecrated in the early seventh century without any reference to Rome.

The exception is also what made Canterbury different: when it was traditional for the bishop of a specific see to be granted the use of the pallium then this needed to be asked of the pope by the new incumbent of the see. The pallium was neither handed on directly, internally within a see from bishop to bishop as a matter of convention, nor dispatched by the pope following a succession as a matter of course; nor was it even simply granted on papal initiative. The pallium had to be formally requested.Footnote 53 This is apparent throughout Gregory's letters, but it is most unambiguous in his letter of September 597 to Queen Brunhild.Footnote 54 Brunhild was the grandmother of the young kings Theoderic and Theodebert and acted at this period in essence as regent for them. She requested that Pope Gregory grant the pallium informally to Syagrius, one of her episcopal favourites. Gregory, ever the diplomat, was keen not to disappoint one of his key Frankish contacts and allies, but there was a problem. As Gregory noted, ‘you [Brunhild] wanted it to be understood that it [the pallium] was sent not due to your request but from us’.Footnote 55 This was not, however, consistent with the rules relating to palliums. Instead Gregory emphasised that, consistent with ‘prisca consuetudo’ (‘the old tradition’), ‘the honour of the pallium ought only to be given to someone making a strong request for it, as the merits of the case demand’.Footnote 56 For the pope to issue the pallium, therefore, it had to be requested by the bishop in question, usually, as in Brunhild's case, supported by the relevant secular authority.

Consequently, when popes issued a pallium to the bishop of Canterbury, as they did following each episcopal accession to the see after Augustine – since the vestment had effectively become a mark of office of the English metropolitan – we can be certain that this followed a request from the bishop and a trip to Rome by representatives of the English Church who would have brought letters containing that request. The English delegations to Rome, of course, would have notified the popes both of the death of the former bishop of Canterbury and of the name of the successor to the see, but this was not the reason for the visit. The purpose of the journey, and of sending the delegation, was to ask the pope to grant the pallium to the new bishop and to bring it back with them.

Indeed, when one examines the evidence for the trips to Rome and the content of the letters that survive in this context, it becomes clear the journeys are associated with requests for the pallium. The letters that Bede includes from the fourth (624–5) and fifth (633–4) visits specifically state that the granting of the pallium was the result of the trip to Rome, and thus necessarily imply that its request had been the primary reason for travelling in the first place. Obviously, that does not preclude the delegation discussing other matters or undertaking additional tasks while in Rome; even so, without the need to request the pallium arising from the death of the previous bishop of Canterbury and the choice of a new one, a party would not have been sent to Rome in the first place.

Given that the letters relating to the third visit (619–20) were also issued in the context of episcopal succession, this journey too must have been occasioned by the need to request the pallium. Indeed, in this light, the same can be said for the first trip considered above. Although during the first visit (600–1), there was no need to notify the pope of a bishop's death, given the plans to raise Canterbury to metropolitan status, Æthelberht and Augustine had decided to request that the pallium be given to the latter as a mark of honour and authority, and thus define the English Church more distinctly as an ecclesiastical ‘province’, more separate from Gaul and the Gallic Church than previously it may have appeared.

In 600, therefore, in line with usual practice, Augustine sent letters to Gregory requesting that the pope bestow the pallium on him. Of course, as Gregory's letters to Frankish recipients reveal, this was not the only request that Augustine made of Gregory. He also asked for reinforcements.Footnote 57 Furthermore the Libellus responsionum suggests that Augustine had some rather more specific questions for his mentor, particularly relating to pastoral, ritual, administrative and disciplinary best practices. Even so, the request for the pallium in 600–1 was central to the rationale for Laurence and Peter's journey to Rome. Almost certainly Æthelberht had added his own voice to Augustine's appeal, sending a letter to the pontiff in support of the request for the pallium, just as Brunhild had done for Syagrius. A similar royal request is explicitly referred to in Pope Honorius' letter to King Edwin in HE 2.17,Footnote 58 and is implicit in that of Boniface to Justus in HE 2.8, where the pope refers to the praise that Eadbald had heaped upon the bishop, thus revealing that he had received letters from the king which surely included support for Justus' request for the pallium.Footnote 59

To summarise: for four of the five delegations that the early English Church sent to Rome for which we have evidence of their purpose, the reason for making the journey was to request the pallium. The visit in 600–1 and the letters relating to it were focused on acquiring the pallium. The same was true for the letters and visits of 619–20, 624–5 and 633–4 (the third, fourth and fifth visits). In each of the latter three cases the pallium was being requested following the death of the bishop of Canterbury and the choice of his successor, something that was obviously unnecessary in 600–1.

Mellitus' visit to Rome, 609–10

The pattern established in examining visits 1, 3, 4 and 5 should now be clear. On the death of a bishop of Canterbury, once a successor had been chosen, representatives of the English Church travelled to Rome with letters asking for the pallium to be given to the new bishop. In each case these letters would have included epistles from the new bishop, and probably also from the king, in support of the request.

This therefore leaves unexplained only the second visit, that of Mellitus in 609–10. Why did Mellitus travel to Rome at that time? The reason Bede provides is scarcely convincing, or at least unlikely to be comprehensive: ‘About this time Mellitus, bishop of London, went to Rome to confer with Pope Boniface about the needs of the English Church.’Footnote 60 Bede's statements about motivations are not to be relied upon: they reflect his best assessment, and, though ultimately only guesswork, are usually unsurprisingly sensible. Nevertheless, because he presents them as simple matters of fact they have a tendency to mislead readers into thinking that he possessed separate information on which to base them, when this is very rarely the case.Footnote 61 Visit ii is a consummate example of this. Bede has no source behind his extremely generic claim. He has simply tried to make sense of why Mellitus might have gone and what he might have done while there; but conferring ‘about the needs of the English Church’ does not amount to an evidenced explanation of Mellitus' purposes in travelling.

Bede at least did not fall into the trap that some modern historians have of believing that because Mellitus attended a synod while in Rome, this was his reason for being in the city in the first place.Footnote 62 These were Roman synods attended by Italian clerics within the Roman province; they were not intended for all ecclesiastics within the Western Church.Footnote 63 There is no reason to think that Mellitus had travelled to Rome specifically to take part in this synod. Mellitus and the English Church as a whole were not invited. The clergy of far-away Kent would not even have known that the synod was happening. Their members would not be expected to attend such a council unless they were already in the city: if they were, however, they would be more than welcome, since their presence would add lustre and an international dimension to an otherwise essentially local event. In a similar situation later in the century, when Wilfrid, by chance, happened to be in Rome at the time of the 680 synod which Pope Agatho had called preparatory to the 680–1 ecumenical council, the exiled bishop of York attended and attested the council's decrees.Footnote 64 But plainly Wilfrid was not in Rome in 680 in order to attend the synod, any more than Mellitus was in 610. Almost certainly, therefore, Mellitus attended the synod in February 610 because he was in Rome; he was not in Rome to attend the synod. Had he not happened to have been in Rome at the time, the English Church may well never even have known that the synod had taken place.

So the question remains: why had Mellitus made the journey to Rome? Unfortunately Bede does not include the papal missives that he mentions. These might otherwise have given us direct evidence for the reasons for the trip. Even so, some deductions are possible merely from his references to the letters. Most relevant is that Bede notes that the letters were sent to Laurence, not to Augustine. Therefore, by this stage, Augustine had died and had been succeeded by Laurence; more important, by the time that Pope Boniface wrote the letters the pope must have been made aware of these events. When might that have happened?

In the light of all the other evidence collected above, it is difficult to avoid the conclusion that it was Mellitus, on this very visit, who had informed the pope of Augustine's death and indeed that his trip to Rome had been prompted – just as all the others set out earlier had been – by the need to deliver a formal request for the bestowal of the pallium on the deceased bishop's successor. Thus, the date that Bede provides for the synod that Mellitus, Laurence's ambassador, attended while in Rome, can indirectly reveal the date of Augustine's death.

The synod canons were dated to 27 February 610. Mellitus therefore arrived in the late summer or, at the latest, early autumn preceding that. Given that we know that Augustine died on 26 May this timing would fit perfectly. There was no need for a long – or any – delay while a successor was chosen, since Augustine had already consecrated Laurence for that purpose.Footnote 65 Mellitus thus made his journey to Rome during the summer of 609, and stayed in the city over the wintertime and into the new year, attending the synod in February 610. He then returned to Kent in spring or summer 610 with letters for the English Church, and the pallium for Laurence.Footnote 66

If these steps are accepted then it is possible to conclude with some confidence that Augustine died on 26 May 609. Indeed it is worth considering the implications if he did not die on this date. If Mellitus' trip to Rome in 609–10 was not when the papacy first heard of Augustine's death, then another delegation must have been sent between 604 and 609 which brought the information about Augustine's passing and Laurence's succession and which requested the pallium for the new bishop. Such a visit would also have produced papal letters to the English Church confirming the granting of the vestment and no doubt also dealing with other topics. Where are these letters? Overall Nothelm seems to have been fairly thorough in his trawl through the epistles in the papal scrinium and in his collection of the available missives relevant to the early English Church. That he brought Bede not only letters to the English Church but also several about the English Church, including a number to correspondents in Gaul,Footnote 67 demonstrates this categorically. It is therefore reasonable to assume that Nothelm brought whatever letters he could find in the papal archives relating to Augustine and his successors. That those we know he did bring correlate so closely with the deaths and successions of bishops of Canterbury is more than mere coincidence. All the trips for which we have evidence of purpose were sent to ask the pope for the pallium. Bede's description of visit ii (609–10) contains no internal indication of purpose. Nevertheless, if one is unwilling to accept that this was the occasion upon which the pope was notified of Augustine's death and sent the pallium back for Laurence, then one is compelled to invent another journey to Rome at some stage between 604 and 609 for which there is no evidence and for which Nothelm could find no letters in the scrinium. Furthermore, Mellitus' trip in 609–10 would remain odd and its motivation unclear. There is no need to multiply entities. The simplest explanation, and the one which best fits the evidence and the analogues, is that it was during this visit, otherwise unexplained, that the pallium was requested and the death of Augustine was reported.

Although in a case such as this – as in so much of early medieval history – absolute certainty is obviously not possible, a date of 26 May 609 for Augustine's death fits well with all the available evidence that we have and is the most probable explanation of the sources at our disposal.Footnote 68 Moreover, it should be emphasised that these sources, though often deduced from the content of the HE, are themselves primary sources – papal letters — even if we no longer possess their full texts. Thus, the conclusion about Augustine's date of death is based not on Bede qua Bede, but on an analysis of the primary sources used by Bede. For all his scholarship and natural instincts as a historian, Bede's narrative is only as good as the material with which he worked and Bede is at best a secondary source for most of the events that he narrates. Arguments based on his account which make no attempt to discern between his sources and their relative reliability can have no permanent value. Wherever Bede's version of events does not rely on original evidence his narrative needs to be treated with more circumspection than has been usual. None the less, since Bede at times used primary sources, if scholars can identify and isolate these and build a narrative upon them, it is possible to construct a more reliable version of events in the early seventh century than that in the HE and occasionally even, as here, make some deductions that Bede himself missed.

References

1 Bede, Historia ecclesiastica gentis Anglorum 2.3: Bede's Ecclesiastical history of the English people, ed. and trans. B. Colgrave and R. Mynors (1969), repr. with corrections Oxford 1991. Translations are taken, with minor modifications, from this edition.

2 This range is based on Bede's statements that Augustine consecrated Mellitus and Justus, as bishops of London and Rochester respectively, in 604 (HE 2.3), and that Pope Boniface iv wrote to ‘archbishop’ Laurence in what amounts to ad 610 (HE 2.4).

3 See, for instance, Margaret Deanesly, Augustine of Canterbury, London 1964, 89, and H. Mayr-Harting, ‘Augustine (d. 604)’, ODNB. Others, such as the designers of the modern plaque placed to mark the former site of Augustine's tomb among the ruins of SS Peter and Paul's monastery in Canterbury, state that he died in 605 – the idea being, presumably, that since Bede's reference to the bishop's death comes after his description of the consecration of Mellitus and Justus in 604, one should add an extra year – just to be safe.

4 Letters of Gregory the Great, i. 71, goes further yet, claiming, again with no evidence or argument, that Augustine resigned.

5 As H. Mayr-Harting put it, ‘There is, however, reason to believe that Bede knew very little about the Gregorian mission except what he could learn from those of Gregory's letters with which the priest Nothelm had supplied him’: The coming of Christianity to Anglo-Saxon England, 3rd edn, Philadelphia 1991, 63.

6 The exceptions are the papal letters relating to paschal computation which he derived, with the rest of his computus material, ultimately from Irish sources: Bedae opera de temporibus, ed. C. Jones, Cambridge, Ma 1943, introduction at pp.105–13. These letters are that of pope-elect John to the Irish, which Bede quotes in HE 2.19, and that of Pope Honorius to the Irish, which Bede mentions in the same chapter, but does not include. Finally, there is also the Libellus responsionum, which Bede inserts in full in HE 1.27. This is a complex document, which there is no space to discuss here: suffice it to say for present purposes that Bede did not obtain the text from Rome and Nothelm; it reached him with other canon law materials. The traditional place to begin studies of the Libellus has been the work of Paul Meyvaert, for instance his ‘Bede's text of the Libellus responsionum of Gregory the Great to Augustine of Canterbury’, in P. Clemoes and K. Hughes (eds), England before the Conquest: studies in primary sources presented to Dorothy Whitelock, Cambridge 1971, 15–33. It was Meyvaert who first showed that the Libellus had reached Bede in canonical collections. For a more up-to-date discussion of the Libellus and its presence in early canon law collections which supersedes, and in important ways corrects, Meyvaert's conclusions, see now Michael D. Elliot, ‘Boniface, incest, and the earliest extant version of Pope Gregory i's Libellus responsionum (JE 1843)’, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte. Kanonistische Abteilung c (2014), 62–111.

7 ‘My principal authority and helper in this modest work’: HE, preface. This was not Nothelm's first journey north with information for the Ecclesiastical history. The HE's preface makes plain that on an earlier trip he had brought material that Albinus had collected about the history of the diocese of Canterbury as well as of other regions and sees.

8 To clarify, terms in this article such as ‘early mission fathers’, ‘Augustinian mission’, or the ‘early Church at Canterbury’, and similar phrases, refer to the Church established by Augustine and his companions prior to the consecration of the first native bishop of Canterbury, Deusdedit in 655 (HE 3.20). Different terms are not intended to indicate different entities: variety is purely for the sake of the reader. As should be evident, the same is also true for terms like ‘trip’, ‘visit’, ‘delegation’ and the like.

9 R: 11.45.

10 HE 1.29 (R: 11.39); HE 1.31(R: 11.36); HE 1.32 (R: 11.37). Bede implicitly dates the Libellus (HE 1.27 [R: 11.56a; M: 8.37]) to the same period, but the preface to this ‘letter’, which Bede did not have, makes this unlikely. Again I would like to express my gratitude to Michael D. Elliot for many stimulating discussions of this intriguing yet perplexing text.

11 R: 11.56.

12 R: 11.35. This is probably the basis for Bede's knowledge that Laurence and Peter had been the messengers sent to Rome on this occasion. This detail is only otherwise present in one version of the Libellus responsionum, but not that which Bede quoted (though Meyvaert, without strong grounds, believed he may have known it as well: ‘Bede's text of the Libellus’, 30–1).

13 For instance, either R: 11.41, copies of which were sent to several Frankish bishops including Menas of Toulon and Serenus of Marseilles, or perhaps more probably R: 11.48 to Brunhild. It is only in these letters that Gregory notes that Augustine had specifically requested reinforcements. Bede states this in terms at the start of HE 1.29, even though it was not mentioned in any of the letters that he quotes.

14 Such letters include R: 11.40, to Bishop Aetherius of Lyons, or R: 11.41, to Bishop Aregius of Gap, and a number of others.

15 There are only two examples of letters to Gaul sent in April: R: 3.33 to Dynamius (from April 593) and R: 5.31 to the tenants of estates or farms in Gaul (from April 595). In addition, one more April letter, that to Leander dated April 591 (R: 1.41), would have made the same journey at least as far as Marseilles, before going on to Spain.

16 R: 7.12 to Respecta, abbess of Marseilles (from October 596) and 11.9 (to Conon, abbot of Lerins) and 11.10 (to Serenus, bishop of Marseilles), both from October 600.

17 R: 13.7–9 (M: 13.5–7) and R: 13.11–13 (M: 13.9–11), which were all apparently written and sent at the same time.

18 ‘absortus fuerat fluctibus Italici maris’: HE 2.20.

19 R: 9.227 (M: 9.229a), with Gregory's reply thanking the king for his gifts, while castigating the envoys for their cowardice, at R: 9.228 (M: 9.229b), from August 599. Columbanus too wrote of the difficulties, including sea travel, which had prevented both him and his letters reaching the pope: Sancti Columbani opera, ed. G. S. M. Walker, Dublin 1957, letter 3, pp. 22–3. Although Columbanus says that the sea is ‘non tam … visibilis quam intelligibilis’, this does not mean there was no actual sea which the saint or his epistles had to cross.

20 An instance is when Pope Stephen ii needed to send envoys to Pepin in 756, circumventing the siege of the Lombard king Aistulf: Peter Llewellyn, Rome in the Dark Ages, 2nd edn, London 1993, 213. The emerging preference for the Alpine route is described in David Pelteret, ‘Not all roads lead to Rome’, in F. Tinti (ed.), England and Rome in the early Middle Ages: pilgrimage, art and politics, Turnhout 2014, 17–41 at pp. 22–31, and Stephen Matthews, The road to Rome: travel and travellers between England and Italy in the Anglo-Saxon centuries, Oxford 2007, 7–11, 39–44.

21 For instance, R: 11.22 to Paschasius, bishop of Naples, or R: 11.25, to John, bishop of Syracuse, both from February 601.

22 A point well emphasised in Pelteret, ‘Not all roads’, 17.

23 HE 1.2.

24 HE 1.17.

25 This is obvious, but – given the lack of contemporary sources for the logistics of the journey – unprovable for the period under consideration. William the Conqueror's intended summer invasion and the unexpected delay which undermined Harold's best-laid plans is one later, if dramatic, piece of evidence for the truism.

26 There is a summary of some of the distances involved in journeys to and from Rome in the Anglo-Saxon period in Matthews, Road to Rome, 8–11. All but one of the trips that Matthews mentions took place over two calendar years. The exception is that of Archbishop Robert in 1051 where there was a specific reason for haste – his need to return to England with the pallium as quickly as possible to show his legitimacy and authority. It should be noted that since Matthews's work is focused almost entirely on a period when the land route through Italy was preferred, the climactic factors affecting sea travel between Gaul and Italy mean that in the early seventh century there were more, and different, limits on the practicable periods of travel than those for which he makes allowance.

27 This does not include that to Mellitus. The fact that Gregory addressed his 18 July letter to Mellitus in Gaul, even though almost a month had passed since he had written the others that Mellitus and company had taken with them, shows that the party was not expected to be travelling particularly rapidly through France on their way back to England.

28 The pallium was a highly revered ecclesiastical garment granted in certain circumstances to senior bishops, such as the bishops of Arles, Ravenna and Milan. Around 600 the vestment was bestowed only on a very few and was allowed to be worn solely under specific conditions. During the seventh century, however, partly thanks to trends emerging from the English Church, especially under Theodore, the pallium gradually became associated in the West with metropolitan bishops more generally, at the same time as the latter increasingly took on the title of archbishop, again apparently following Theodore-inspired English influence. Helpful summaries and simplifications of some of the issues surrounding palliums can be found in A. Thacker, ‘Gallic or Greek? Archbishops in England from Theodore to Ecgberht’, in P. Fouracre and D. Ganz (eds), Frankland: the Franks and the world of the early Middle Ages: essays in honour of Dame Jinty Nelson, Manchester 2008, 44–69 at pp. 48–54, and Story, Joanna, ‘Bede, Willibrord and the letters of Pope Honorius i on the genesis of the archbishopric of York’, EHR cxxvii (2012), 783–818 at pp. 790–1Google Scholar.

29 ‘Deo dilecto archiepiscopo Laurentio et clero uniuerso, similiter et Aedilbercto regi atque genti Anglorum’: HE 2.4. The term ‘archbishop’ has been added anachronistically by Bede. Theodore was the first bishop of Canterbury to use this title: for a discussion on this point see T. M. Charles-Edwards, Early medieval Ireland, Cambridge 2000, 416–21, and Thacker, ‘Gallic or Greek?’, 55.

30 ‘anno octauo imperii Focatis principis, indictione xiii, tertio die kalendarum Martiarum’: HE 2.4.

31 In this section Bede may, however, have adapted some of his language from that in the papal letter: for instance in the phrase describing Mellitus and Justus as those ‘Qui cum magna ecclesiam Anglorum cura ac labore gubernarent’(‘guiding the English Church with great care and energy’).

32 It is arguable that rather than referring to letters sent at the same time to Mellitus as bishop of Canterbury and to Justus as bishop of Rochester, Bede intended here (2.7) to summarise the letters that Pope Boniface wrote during his pontificate to the metropolitans of the English Church. According to such an interpretation, the ‘scripta exhortatoria’ to Mellitus and Justus mentioned in this chapter would include both a letter to Mellitus as bishop of Canterbury sent following Laurence's death and the later letter (which Bede included in HE 2.8) to Justus once he in turn became bishop of Canterbury. I am grateful to Alexander Callander Murray for this suggestion.

33 Bedae opera historica, ed. C. Plummer, Oxford 1896, ii. 90.

34 Bede says that Mellitus and Justus, ‘susceperunt scripta exhortatoria a pontifice Romane et apostolicae sedis Bonifatio, qui post Deusdedit ecclesiae praefuit, anno incarnationis dominicae DCXVIIII’ (‘received letters of exhortation from Rome from Pope Boniface, who succeeded Deusdedit in the year of our Lord 619’): HE 2.7.

35 See, for instance, R. Sharpe, ‘King Ceadwalla's Roman epitaph’, in A. Orchard and K. O'Brien O'Keefe (ed.), Latin learning and English lore: studies in Anglo-Saxon literature for Michael Lapidge, Toronto 2005, 171–93, and Lapidge, M., ‘The career of Aldhelm’, Anglo-Saxon England xxxvi (2007), 15–69CrossRefGoogle Scholar.

36 Thus the claim inserted into some manuscripts of the Liber pontificalis biography of Gregory's predecessor Pelagius ii that he died in the fifth indiction would give a date of death of 587, rather than 590, the correct year: Liber pontificalis, ed. L. Duchesne, Paris 1886–92, i. 309.

37 HE 2.1.

38 In fact they probably had two: both in terms of indictions and imperial years.

39 ‘Scripta quae perlatores ad sanctae memoriae Seuerinum papam adduxerunt, eo de hac luce migrante, reciproca response ad ea, quae postulate fuerant, siluerunt’ (‘The writings which were brought by envoys to Pope Severinus of holy memory, were left with the questions contained in them unanswered when he departed this life’): HE 2.19.

40 The LP omits mention of the full length of the vacancy between Deusdedit and Boniface, which was actually over a year, not just one month and sixteen days. Therein may lie the explanation for Bede's faulty conclusion about Gregory the Great's incarnational year of death. Working back from 619 as Boniface's year of accession, inferred, as argued above, from the 620 letters that he possessed, and using the LP 's figures for vacancies and length of pontificate in the Roman see, the missing year in that text would mean that Bede arrived at 605 for Gregory's death, rather than 604.

41 ‘Susceptis namque apicibus filii nostril Adulualdi regis repperimus, quanta sacri eloquii erudition eius animum ad uerae conuersionis et indubitatae fidei credulitatem fraternitas uestra perduxerit’ (‘We have learnt from the letters received from our son King Eadbald how you, brother, by your learning and holy eloquence have guided his soul to the assurance of true conversion and a state of real faith’): HE 2.8.

42 This is especially so in regard to the reference to Eadbald's conversio by Justus; assuming, as I think we must, that the Aduluald that Justus is said to have converted in Bede's version of the letter to Justus should be equated with the Audubald that Justus is said to have converted in the letters to Edwin (HE 2.10) and to Æthelburh (HE 2.11) and that both Aduluald and Audubald refer to King Eadbald of Kent.

43 This may be, as Meyvaert thought, because the letters, in the form in which Bede possessed them, had already lost their dating clauses: ‘The Registrum of Gregory the Great and Bede’, Revue Benedictine lxxx (1970), 162–6. But it is just as likely, if not more so, that Bede removed the dates for the sake of simplicity because their content complicated, if not contradicted, the narrative of the Northumbrian conversion and particularly the position of Paulinus that Bede had heard or reconstructed prior to receiving the papal letters from Nothelm.

44 Unless, perhaps, Bede's statement that Mellitus died in 624 at the end of HE 2.7 comes from information in these letters of Boniface which probably date to 625.

45 HE 2.7.

46 In the letter quoted in HE 2.8.

47 This can be reconciled with Bede's date for Justus' consecration of Paulinus as bishop (21 July 625), although, despite what Bede implies, possession of the pallium was not necessary for Justus to consecrate other bishops. The precise chronology of Edwin's conversion though is a more complex question beyond the scope of this paper. If Bede is correct in saying that Bishop Romanus of Rochester's death in the Ligurian Sea occurred on a mission for Justus (HE 2.20), then it was probably on this journey. But Bede's mention of Justus' role may be no more than an assumption on his behalf; in which case Romanus is more likely to have drowned during visit v (633–4).

48 ‘Data die III id. Iun. Imperantibus dominis nostris piissimis Augustis Heraclio anno XXIIII, post consulatum eiusdem anno XXIII, atque Constantino filio ipsius anno XXIII et consulatus eius anno III, sed et Heraclio felicissimo Caesare idem filio eius anno III, indictione VII’: HE 2.18. Pope Honorius’ letter to King Edwin, which Bede inserted in HE 2.17, was no doubt sent at the same time, although the version included in the HE does not have a dating formula.

49 ‘id est, anno dominicae incarnationis DCXXXIIII’: HE 2.18.

50 ‘Ea uero, quae a nobis pro uestris sacerdotibus ordinanda sperastis, … gratuito animo adtribuere ulla sine dilatione praeuidemus’ (‘We are preparing to concede you willingly and without delay those rights which you hoped we should grant your bishops’): HE 2.17 (to Edwin); ‘iuxta uestram petitionem quam filiorum nostrorum regum’ (‘in accordance with your request and that of the kings our sons’): HE 2.18 (to Honorius). It should be noted that the contents of the letter to Bishop Honorius (HE 2.18) show that Pope Honorius' original missive was intended for Bishop Honorius and Bishop Paulinus as joint addressees with a copy to be sent to each of them.

51 HE 2.20.

52 Romanus' death may well have occurred on this mission, but since Bede says that he was drowned on an embassy for Justus, he may have led the delegation in 624–5, immediately after his consecration as bishop of Rochester.

53 Later it had to be requested by the relevant bishop himself in person; but that was not the case in the period under consideration here.

54 R: 8.4 (M: 8.1).

55 ‘Deinde quod non id ex vestra petitione, sed ex nobis transmissum voluistis intellegi’: ibid.

56 ‘honor pallii nisi exigentibus causarum meritis et fortiter postulanti dari non debeat’: ibid.

57 As Gregory put it in R: 11.41 (dated 22 June 601), copies of which were sent to various Frankish bishops including Menas of Toulon and Serenus of Marseilles, ‘Augustinus eos qui secum sunt ad hoc opus exequendum per diversa loca asserat non posse sufficere’ (‘Augustine asserts that those who are with him are insufficient to carry out this work through the various locations’). Therefore, as the pope explained in R: 11.48, a letter written the same day to Queen Brunhild, he was sending more monks, ‘quod illos qui secum sunt sufficere sibi dicit non posse’ (‘due to the fact that he [Augustine] says those with him cannot be sufficient for him’).

58 ‘Ea uero, quae a nobis pro uestris sacerdotibus ordinanda sperastis, hoc pro fidei uestrae sinceritate, quae nobis multimoda relatione per praesentium portitores laudabiliter insinuata est, gratuito animo adtribuere ulla sine dilatione praeuidemus; et duo pallia utrorumque metropolitanorum, id est Honorio et Paulino, direximus’ (‘We are preparing to concede you willingly and without delay those rights which you hoped we should grant your bishops: we do this on account of the sincerity of your faith which has been abundantly declared to us in terms of praise by the bearers of this letter; and so we are sending a pallium for each of the two metropolitans, that is for Honorius and Paulinus’): HE 2.17. The royal request in this case is also mentioned in the letter to Bishop Honorius: ‘iuxta uestram petitionem quam filiorum nostrorum regum’ (‘in accordance with your request and that of the kings our sons’: HE 2.18).

59 HE 2.8.

60 ‘His temporibus uenit Mellitus Lundoniae episcopus Romam, de necessariis ecclesiae Anglorum cum apostolico papa Bonifatio tractaturus’: HE 2.4.

61 As James Campbell noted, Bede ‘did not write in an age in which it was thought necessary to distinguish between known facts and deductions or assumptions’: ‘Bede I’, in James Campbell (ed.), Essays in Anglo-Saxon history, London 1986, 1–27 at p. 9.

62 For instance, N. Brooks, ‘Mellitus (d. 624)’, ODNB.

63 For a summary of the characteristics of Roman synods in the seventh century see Llewellyn, Rome in the Dark Ages, 129–30.

64 H. Chadwick, ‘The English Church and the monothelite controversy’, in M. Lapidge (ed.), Archbishop Theodore: commemorative studies on his life and influence, Cambridge 1995, 88–96 at p. 92.

65 HE 2.4.

66 Even though Bede possessed copies of these papal letters it is unlikely that he realised that they included sufficient information to deduce the date of Augustine's death. Pope Boniface would not have mentioned that Augustine had died. None of the comparable papal missives, for instance to Justus or Honorius (included in HE 2.8 and 2.18), told the recipient of the death of his predecessor. Such a detail would be completely superfluous, especially as the epistle might be received perhaps as much as a year and a half after the last bishop of Canterbury's death. Boniface's letters would have mentioned the pallium; but it should not surprise us that Bede does not draw attention to this element in the text in front of him. Bede took it for granted that bishops of Canterbury would be granted the pallium: this was simply standard practice requiring no additional comment or emphasis. Bede only laid emphasis on the pallium in the HE for good reason, which for him was always the same reason. He only really focused on the pallium when it related to Northumbria: in HE 1.29 when Gregory told Augustine there should be a metropolitan see at York; in HE 2.8 when the gift of the pallium to Justus is underlined because Bede (wrongly) believed that only thanks to the possession of the pallium could Justus consecrate Paulinus for York, as well as Romanus for Rochester; and in HE 2.17 and HE 2.18 where the pallium was sent to Paulinus as part of the creation of the Northumbrian ‘province’ centred on York. Bede also mentioned the pallium once outside a papal letter context, but again with a Northumbrian connection, when he noted in HE 2.20 that going into exile on Edwin's death and becoming bishop of Rochester, Paulinus left his pallium in that see at his passing.

67 Examples are those that Bede included in the HE at 1.24 and 1.28.

68 A date of death for Augustine might be thought to provide a date of death for Peter, abbot of SS Peter and Paul, Canterbury, the miracles surrounding whose tomb following his death at sea are described by Bede in HE 1.33. As the updated ODNB online entry for Peter concludes: ‘The year of his death probably depends on that assigned to Augustine, since Thomas Elmham, the early fifteenth-century chronicler of St Augustine's (as St Peter's and St Paul's became in 978), states that Petrus died one year, seven months, and three weeks after Augustine’: William Hunt, ‘Petrus [St Petrus] (d. 605 x 11)’, rev. Marios Costambeys, ODNB. But in reality Thomas of Elmham is no more reliable as a source for this date than he is for any early event in Canterbury's history. Independent primary source evidence shows that such a ‘tradition’ is definitely flawed since Peter attested the canons of the 614 Council of Paris: Les Canons des conciles mérovingiens (VIe–VIIe siècles), ed. C. de Clerq and trans. Jean Gaudemet and Brigitte Basdevant, Paris 1989, ii. 524–5.