To complicate matters still further, there were other ways of calculating the year. In the Middle Ages each new year began according to the local customary practice or the particular allegiance of the person computing it. In England, the financial and legal years were divided into terms, Michaelmas, Hilary, Easter and Trinity, the new year beginning with Michaelmas on 6 October. These cut across the most popular form of reckoning, which was the regnal year, dating from the start of a new king’s reign. Regnal years were used by those in the employment of popes and bishops, kings and princes, and therefore, of necessity, varied from region to region. The regnal year of Henry V, for example, began on 21 March 1413, the first full day after his father’s death.
12
Throw into this equation the fact that most dates were not given as simple consecutive numbers and that there was an uneasy mismatch between the Christian and Julian calendars, which were both in use at the same time, and one can begin to see why historians and chronologers in the Middle Ages sometimes made mistakes. In the Christian calendar, dates were referred to as the names of Church festivals and saints’ days, including not only the day itself but the day before (
pridie
and
vigilia
) and “the day after” (
crastinum
). In the Julian calendar, every month was unequally divided into the periods of calends, nones and ides, within which the days were counted in numerically descending order. According to this system, our 25 October was thus the eighth day before the calends of November, whereas 30 October was only the third.
A medieval writer wishing to give the date of the battle of Agincourt would have several options before him. Chivalric authors tended to go for the easy option: Monstrelet, for example, simply called it “Friday, the xxvth day of the month of October, one thousand four hundred and fifteen.”
13
Ecclesiastical writers, including chroniclers and clerks in the royal administration, perhaps because they were more numerate and more obligated to abide by Church practices, used the more complicated systems. A Church-trained English and French chronicler would have described the same event in different terms. Neither would have referred to it as being fought on 25 October, but on the Feast of St Crispin and St Crispinian. The Englishman might have placed it “in the third year of our lord king Henry the fifth of that name after the conquest.” His French counterpart, writing in the name of Charles VI, would have described it as being “on the Feast of St Crispin and St Crispinian, in the thirty-fifth year of our reign.” That was why every medieval chronicler and clerk had to have a set of chronological tables to hand whenever computing a date.
It was not even easy to tell exactly what time it was. Although it was generally accepted that there were twenty-four hours in a day, how these hours were measured varied. There were three systems in use in the early fifteenth century. One was the early medieval custom of splitting the day into two periods, from sunrise to sunset, and sunset to sunrise, each of which was artificially subdivided into twelve unequal hours. In winter, the daylight hours would be shorter and the night-time ones longer, a situation that reversed in summer. The second method also varied according to the season, and was determined by the seven canonical hours that marked the principal daily services in the church; they began with Prime at daybreak and ended with Vespers as darkness fell. The advantage of this system was that, although it varied from place to place because it depended on the time that the sun rose and the hours were again of uneven length, the services were marked by the ringing of bells in monasteries and parish churches, which were audible to the people living around them. Like the schoolbells and factory horns of the modern world, these determined the length of the working day for the vast majority of people.
14
The third method of telling the time, which was only reluctantly adopted in some monastic houses, was entirely divorced from the seasons. Mechanical clocks divided the day into twenty-four hours of equal length and measured from midnight to midnight. Sundials and hourglasses, filled with water or sand, had been in use for centuries, but the new clocks were made out of precision-crafted moving parts of iron. The earliest recorded example in England was made in 1283 by the canons of Dunstable Priory, but the oldest surviving one, in Salisbury Cathedral, dates from a century later. Many of these clocks were works of remarkable craftsmanship: in 1322 the priory of Norwich Cathedral had one with a large astronomical dial and automata, including fifty-nine images and a procession of monks. By the fifteenth century, mechanical clocks dictated the time in most Benedictine houses and were displayed on churches and other buildings for the benefit of the wider population.
15
Our poor chaplain, trying to work out the date of the king’s departure from Harfleur, had to struggle with all the conflicting elements of the medieval calendar. In his praiseworthy efforts to be precise, he merely muddied the waters. They had set off, he decided, “on Tuesday, the day before the feast of St Denys, on the nones of October.” The feast of St Denis was 9 October, and in 1415 it fell on a Wednesday, so the day before would indeed have been a Tuesday. Unfortunately, the nones of October, according to the classical Roman calendar, were on 7 October. It is likely that this was just a slip of the pen, or of the finger as he cross-checked his chronological tables.
16
What is more difficult to explain is why the chaplain thought they left Harfleur on Tuesday 8 October, when the exchequer records explicitly state that it was on Sunday 6 October. There is no obvious answer to this question, but it is likely that the exchequer arbitrarily selected 6 October for reasons of administrative convenience, that being the first day of the new financial quarter for the campaign. All that can be said for certain is that the chaplain was there in person and that, in view of his profession, he would surely have known if they had set off on a Sunday. On this admittedly slim basis for a decision, we shall follow the chaplain, but bearing in mind that he may have been a couple of days out on his reckoning.
17
On Tuesday 8 October, therefore, the king, with his nine hundred men-at-arms, five thousand archers and numberless assorted civilians, including the royal surgeons, minstrels, heralds and chaplains, set out from Harfleur by the Montivilliers road. As was the customary practice, the army had been divided into the three battles, or divisions, in which it would fight. The honour of leading the vanguard, or first division, fell once more to the indomitable Sir John Cornewaille and Sir Gilbert Umfraville. The main body of the army was led by the king himself, with some of his younger and less experienced noblemen, including his twenty-four-year-old brother Humphrey, duke of Gloucester, twenty-year-old Sir John Holland, who had distinguished himself at the landing and during the siege in the company of his stepfather, and John, Lord de Roos, who had inherited his father’s estates the previous year and was still only eighteen or nineteen. The leadership of the rearguard, like that of the vanguard, was entrusted to seasoned campaigners, in this case the veteran Edward, duke of York, and Richard de Vere, earl of Oxford.
18
Within the three divisions, it is likely that the men were still grouped according to the retinues into which they had originally been recruited. There was, in other words, no separating out of the men-at-arms from the archers, even though the latter now outnumbered the former by more than five to one, instead of the usual ratio of three to one favored by the English. What must have taken place, however, was a considerable amount of reorganisation. Many retinues, including Clarence’s, which, with almost a thousand men, had been the single largest company at the beginning of the campaign, had lost their leaders. Many more had seen their numbers reduced by as much as a third. In order to maintain the command structure and discipline, it was important that new leaders should be appointed. In some cases this meant that someone from within the retinue took over, as Sir Thomas Rokeby did when the earl marshal was invalided home. In others, particularly where large numbers of archers were involved, the men would be reassigned to other retinues to restore their numbers to a fighting unit.
19
Given the distance that the army had to travel and the possibility that it would soon have to face battle, it is likely that most if not all the men were mounted. There were plenty of surplus horses at Harfleur, for the priority had been to ship home the men who were sick, rather than their mounts. Together with all the spare horses that everyone above the rank of archer was permitted to take, and the packhorses required to carry the baggage, there must have been a minimum of twelve thousand horses in the column, and it may well have been double that figure.
20
Though the horses were essential for the army’s speed and mobility, their presence in such large numbers meant that it would be difficult to keep them all adequately fed and watered on the journey.
The English could not, and did not, expect their march to be unopposed. On 7 October William Bardolf, the acting lieutenant of Calais, wrote to John, duke of Bedford, Henry’s lieutenant in England, warning him that he had heard reports from both France and Flanders that “without fail” the king would have battle against his adversaries within fifteen days at the latest. Around five thousand Frenchmen had already assembled and “a notable knight,” with a force of five hundred men, had also been posted to the defence of the French frontier against the Calais marches. A reluctant bearer of bad tidings, Bardolf excused himself by explaining that “I thought I ought to tell you this.”
21
Bardolf’s reports were entirely accurate. Ever since 2 August, when the truces between England and France in this area had elapsed, the Calais garrison had launched a number of diversionary raids to distract the French away from Harfleur during the invasion and siege. David, sire de Rambures, grand-master of the crossbowmen of France and captain of Boulogne, had been sent to defend the area from the depredations of the garrison. Despite anguished messages from the townsmen of Boulogne, who sought him out at Fécamp and Rouen, it was not until Harfleur was on the very point of surrender that de Rambures at last obtained permission to send the sire de Laurois, commander of Ardres, with a force of five hundred men, to garrison Boulogne.
22
The arrival of de Laurois and his men was a serious setback to Henry’s plans. He had intended that a three-hundred-strong force from Calais should be sent to Blanche Taque to secure the river crossing of the Somme in readiness for his own arrival. The heightened state of security around the Pas-de-Calais would now make it nearly impossible for this expeditionary force to reach the Somme. When it eventually set out, it was ambushed and overwhelmed by a band of Picards, who killed some of the men and took the rest prisoner, intending to hold them to ransom. The failure of this venture would have important consequences for the English army’s attempts to cross the Somme, the single greatest barrier between them and the safety of Calais.
23
While Henry was besieging Harfleur, Charles d’Albret and Marshal Boucicaut had anticipated his next move by destroying bridges and breaking causeways across all the major rivers and reinforcing and resupplying the towns and castle garrisons throughout Normandy and Picardy. The preparations at Boulogne, which was close to the Pas-de-Calais and therefore faced a double threat of a combined assault from the king in the south-west and the Calais garrison in the north-east, were typical. Strict orders had been given as early as 15 September that the nightly watches were to be augmented with dogs and lanterns set beyond the moat; during the day, watchmen were also to be stationed on the hills above the town, to give early notice of an English approach. The embrasures in the town walls and towers were widened to facilitate crossbow fire, miners were hired and the suburbs were cleared away in anticipation of a siege. David, sire de Rambures, provided a “long bombarde” which was mounted on the walls and the town’s own cannon was brought up from the guildhall. Gunners were brought in to man the artillery and supplies of saltpetre and the other ingredients required for making gunpowder were bought from St Omer.
24
There was little that could be done to protect the country people, but this was an area that had been invaded and had suffered the depredations of war so many times that its inhabitants had long ago learnt that their safety depended on their ability to disappear into local forests and caves. In some instances, caves provided a remarkably sophisticated refuge. At Naours,
25
just to the north of Amiens, subterranean chalk quarries had been artificially enlarged and used as places of safety for centuries. Working along the seams of chalk, which were sandwiched between layers of impenetrable silex, an underground city had been created which was capable of sheltering up to two thousand people at once, together with their sheep, cattle, horses and mules. Twenty-eight galleries led to three hundred chambers, each one large enough to house a family of eight, and to a number of public rooms, including a chapel, a law court and a jail. Excavated at three different levels, between 100 and 140 feet below ground, the cave system was naturally dry, enjoyed a constant temperature of 48 degrees Fahrenheit and had access to the river for water. Six chimneys provided ventilation and enabled food to be cooked. So that the smoke did not betray the presence of the people hiding below, the outlets were over 130 feet away and two of the chimneys vented into the local millers’ houses on top of the hill, giving the impression that the smoke came from their own domestic fireplaces. Any intruders who stumbled upon the entrances found themselves lost in a maze of narrow winding corridors, or ambushed when bending double below doorways deliberately set too low.
So well hidden and secure was this underground city that it was in constant use from Roman times until the end of the seventeenth century. Rediscovered in 1887, after a lapse of almost two hundred years, it found a new lease of life in the bloodbath of the twentieth century, serving as the headquarters of English, Canadian and Australian troops in the First World War and of Rommel in the Second. Nevertheless, early fifteenth-century graffiti and coins dating from the reign of Charles VI, which were discovered in the lowest-level chambers and galleries, indicate that the people of the Naours district also fled there in response to Henry V’s invasions of Normandy.