Otford & District Archaeological Group (ODAG)

The Romano-British Cremation Cemetery at Frog Farm, Otford, Kent, in the context
of contemporary funerary practices in South-East England by Clifford P. Ward 1990

Romano-British Funerary Practices in the 1st and 2nd Centuries A.D.

Hominem mortuum in urbe ne sepelito neve uxito

Thou shalt not bury or burn a dead man within a city
                                                                        - regulation of the Twelve Tables
                                                                reinacted by Antoninus Pius (138 - 161).

As a result of the regulation of the Twelve Tables the vast majority of citizens of the Roman Empire were buried outside the settlements to which they belonged, often grouped along the roads leading from them, where their monuments would recall the dead to the minds of the living (Collingwood & Richmond 1969). The grandiose tombs of the environs of Rome were paralleled in many towns and cities of the Empire, and on a reduced scale in the vicinity of lesser settlements throughout the Roman World, including Britain. In this tradition, a cemetery was established on the ancient road which runs along the lower slopes of the North Downs, now known as the Pilgrims Way, some 300 yards (275 metres) east-south-east of a Romano-British settlement, and 1250 yards (1150 metres) from the nearest definitely located Roman Villa, east of the River Darent, with another 500 yards (450 metres) postulated to the south. The settlement flourished throughout the Roman period, and may well have acted as a market for the surrounding rural area, without developing into a town, the inference being that administration was carried out from the town of Durobrivae (Rochester) some sixteen miles distant along the Pilgrims Way (Rivet 1964). It is likely that Otford supplemented the other minor settlements in the vicinity as a distribution centre as well as acting as a minor centre for iron-working (Pyke forthcoming) and pottery production of Patchgrove wares (Breen 1987).
   There appears to have been a fundamental difference in attitude towards the after-life between the Romans and Celts, the former looking without much enthusiasm on the prospect, at least until the 2nd. century A.D., with 'sit tibi terra levis’ (let the earth lie lightly upon you) being the recurrent theme. An equally pessimistic epitaph on a Wroxeter tombstone proclaimed, "drink while your star gives you time for life", and that on a memorial stone to a thirteen year-old York girl refers to 'the meagre ashes and the shade, empty semblance of the body seeking the manes who dwell in Acheron, the river of woe (Henig 1984).
   The care shown by the Romans in their burial practices was to ensure that the requisite minimum covering of earth prevented the deceased from haunting the living, the fear of ghosts being to the forefront of men's minds (MacDonald 1977). Toynbee considered that the denials of survival of the Epicurean and Stoic systems were not universally accepted by the Roman people, and that human life was not invariably regarded as merely "an interlude of being between nothingness and nothingness", there being some conviction of a continuing soul whereby the living and dead can affect each other mutually. This more optimistic approach gained momentum from the spread of eastern-cults, with the ultimate belief in the resurrection of the body inspired by Christianity, which found official favour in the early fourth century (Toynbee 1971).
   The Celts on the other hand, had a long-held "strong belief in a positive and tangible afterlife where the accoutrements of earthly life will be required and the terrestrial status retained" (Green 1986).
The Roman last rites in the Republican period comprised both inhumation and cremation (Toynbee 1971), but from the early Imperial age that of cremation became almost exclusively used, until a growing belief in the survival of the body pervaded pagan belief, closely followed by Christianity which gained credence throughout the Empire, especially after the official toleration, granted by Constantine in the Edict of Milan in A.D.313 (Somerset Fry 1984).
   In the pre-conquest Iron Age in Britain, there was doubtless ceremony attendant on death, as in the preceding Bronze and Stone ages, but, comparatively little tangible evidence remains. Caesar comments in Book VI of de bello Gallico on the opulence surrounding Gallic aristocratic funerary rites, but makes no reference to ultimate disposal of the remains. Of course, regional and or religious belief variation may have occurred to some degree, but it is likely that, in a tribal society as Iron Age Britain, uniformity would be found within the tribal spheres of influence, and it is likely that many of the ruling aristocracies had links in a common ancestry; their religious beliefs are likely to have been maintained, although the requirements of social custom may have been as strong a factor at the time in regulating the funerary rites (Henig 1984), while the act of passing into the After-world, an occurrence familiar to everyone involved in communal life, must have had considerable influence on the thought-processes and superstitions of the community (Kendall 1982). Darvill identified a Core Zone, comprising the lower Thames valley, including Kent, Surrey, Sussex, London, Essex, Hertfordshire, and parts of Buckinghamshire and north Hampshire which by the mid-1st century B.C. enjoyed a superior economy and close links with the continent, and, in common with the rest of the area, in Kent cremation seems to have been fairly well-established from c. 40 B.C. (Darvill 1987) where cremation cemeteries, although not exclusive, are ubiquitous in utilising small pits to hold ashes placed in urns. These are known from their type-sites of Aylesford and Swarling. Often the ashes were placed in wooden buckets, accompanied by rich grave goods. Other graves contained from many to few, or no, grave goods, and are considered to be of Belgic tradition.
   There is a discernible trend in the later Iron Age towards inhumation in Sussex, which, separated by the relatively uninhabited Weald, could develop a separate non-Belgic tradition (Bedwin 1978). The development of the wealden iron industry in the Roman period has not to date provided any tangible evidence of burial rites (Money 1990), but cremation seems to have been current in the earlier phase of Roman Chichester (Down 1978).
   Recent researches indicate that there was a major development towards a more settled form of land-ownership in the late pre-Roman Iron Age in lowland Britain, and that the Roman Conquest accelerated rather than initiated the process (Salway 1988), it being generally accepted that there was a considerable increase in the numbers of country dwellers in the Roman period (Bird 1987). Investigations suggest that in Iron Age society only a minute proportion of the population was accorded burial rites, probably the elite (Green 1986), but after the conquest it appears that both high and low were afforded burial or cremation, though with differing elaboration.
   The imposition of Roman rule in the south of England in A.D.43 probably had little impact on the everyday lives of the ordinary people, and beyond the final crushing of the British druidic theocracy in Anglesey in A.D.60 by Suetonius Paullinus (Frere 1973) possibly in response to its human sacrificial practices, barbaric, even by Roman standards, or, as hinted by Pliny the elder, through their morale-lowering gloomy prophesies (Somerset Fry 1984), their religious beliefs were probably allowed to continue unhindered, with a gradual syncretism of Roman and Celtic dieties (Clayton 1980). The part played by the Druids in Celtic society is still unclear, as their ritual and theology was unwritten, but they probably were socially, rather than politically, very powerful. They appear to have had their origins in Britain, and after the Anglesey rout they do not reappear in Roman Britain. However, they were active in the Gaulish uprising of A.D. 69-70, and druidic prophesies against the emperors are recorded in the third century. They also continued to wield influence in Ireland as magicians and soothsayers.
   The Roman authorities were in the main extremely tolerant towards foreign religions, provided they were not redolent of political conspiracy or repugnant rites, and they tended to conflate local deities with the Roman pantheon (Salway 1988). Hence as a funerary rite cremation continued to be practised in a more or less traditional manner, and first and second century A.D. examples abound in districts where settlement was effected or continued.
   With a few notable exceptions e.g. St. Pancras, Chichester, and Lankhills, Winchester, relatively few cremation cemeteries have been excavated using modern techniques, and no direct parallels to Otford have been traced. That at Ospringe, which served a still unlocated but substantial nearby settlement, bears many similarities, although it is considerably larger and has inhumations as well as cremations (Whiting et al 1931). Some cemeteries were discovered long before adequate recording techniques were invented, and tantalising glimpses alone remain as of "a Roman cemetery discovered upon East Hill, near Dartford A.D.1792, in which have been found great numbers of urns, intermingled with stone and wood coffins, lachrymatories etc. together with the square pits in which the bodies were burned" (Dunkin 1844). Also near Dartford, another depressing note records that at new gravel pits at Joyce Green, "workmen found several Roman urn burials of the ordinary kind, consisting of small groups of urns here and there" (Payne 1897).
   Other opportunities have been lost even recently, such as the presumed cremation cemetery adjacent to Canterbury Castle where there were "vague reports of further cremations" seen during construction at the Gas Works in the 1950s (Bennett et al 1982).
   A number of deep shaft graves of Iron Age date from South-East England suggest a multiple, possibly family, use, and shaft burials continue into the Roman period. There is much speculation as to the religious significance of these shafts, many of which are classed as 'ritual pits', apparently linking the world of the living with the underworld, and suggestive of propitiation of chthonic deities. These are predominantly, but not wholly, found in areas of Belgic influence, and some, such as South Cadbury, where a young man was buried head-down in a pit associated with alterations to the Iron Age defences, suggest a dedicatory sacrifice (L. Alcock 1972).
   An extremely well-constructed rectangular oak shaft, 12ft. in depth was discovered during railway building at Bekesbourne Hill, near Canterbury, in 1858 at a depth of 13ft. below the modern ground level. It was built of nicely mortised and tied oak baulks and had an internal diameter of 3ft. 3ins. (l metre). At the bottom was a large (?) quern stone on which was placed a circle of horse teeth. Above this were five urns apparently covered with fabric. These in turn were covered by a layer of flints on which stood a further urn again covered by flints. The urns almost certainly contained cremated bones. The structure was roofed with more oak beams and was in remarkably good condition, but was destroyed largely due to the efforts of treasure-seeking navvies (Brent 1859).
   A recently discovered shaft at Deal, Kent, contained a chalk figurine comprising a 'Celtic' head surmounting a dressed block. Associated pottery dates this to the first or second century A.D. and this is a pointer to the idea of communication with the underworld powers (Green 1986). Many shafts contained animal bones, especially of dogs, sometimes with human remains, and the 225ft. deep well at Findon, Sussex, contained inter alia the remains of a horse (personal observation). So-called ritual shafts in or near cemeteries in South-eastern England with rich assemblages of pottery, animal bones and other artefacts, but without human remains, taken as depositories for offerings, have been found (Black 1987).
At Keston, Kent, a shaft l6ft. deep and 11ft, wide was excavated and found to contain cremated remains of two dogs together with pottery described as indeterminate but tentatively ascribed to the 3rd.-4th. Century A.D. (Piercy Fox 1968, Jessup 1970 ), while a well at Staines, Middx., contained parts of no fewer than 17 dog skeletons (Bird 1987).
   Graves, as such, from the Iron Age are very uncommon, and it appears that if dug, they were so shallow as not to have survived later disturbances of the ground. A further possibility is that there was some form of above-ground burial chamber or charnal house, where bodies were either exposed, or, more likely, laid to rest in wooden structures for at least a limited period of time (Black 1987). The ubiquity of the so-called 'four post' structures in Iron Age settlement sites gives weight to such a hypothesis, and the use by the Maoris of New Zealand of such four-post bases for cantilevered rectangular storage huts illustrates the practicality of such (personal observation).
   This would allow the veneration of ancestors by the living, a likely attitude among people who are believed to have placed great store on qualities which could be inherited, even from enemies vanquished in battle, through exposure of severed heads in their dwellings (Delaney 1986). The eventual disposal of the skeletons seems to have been of lesser consequence, as disarticulated bones have been recovered from rubbish pits (L&R Adkins 1983).
   Attempts have been made to identify specifically Roman military cemeteries but this has been largely unsuccessful despite the considerable number of military tombstones known especially from northern Britain, leading to the conclusion that after the conquest the soldiery played a part in developing local community traditions. The early garrisons followed their own ethnic or Roman customs, but gradually the military and civilian traditions mingled and ‘the early Roman style of cremation burial, usually in a pot with an assortment of Romanised goods as grave furnishings, became general for most of the Romanised population’ (Jones 1984).
   It has been established that at a number of sites in Southern England cremation was practised on a fairly large scale, with structures identified as collective crematoria, as at Colchester, Essex, (Hull 1958), and outside Verulamium (Davey 1935), where confined areas of burning some 1.8/2.4m by 0.7/0.9m (5ft. 10ins/6ft. l0ins. by 2ft 3ins./3ft.) in extent, suitable in size for cremation of a body, were located within cemeteries. Toynbee, quoting Roman authors, postulates that the cremation took place either at the place of burial, or at a special crematorium. The wood-pyre was rectangular and the body placed upon it with opened eyes. Various gifts and possessions were added, and sometimes even pet animals were killed as companions to the afterlife. After the mourners had called the deceased by name a final time, the pyre was kindled with torches. The ashes were ultimately drenched with wine, and the ashes and bones were subsequently collected by the relatives and placed within receptacles which varied with the wealth of the family from marble ash-chests or caskets, precious metal vases, to lead, glass or earthenware vessels (Toynbee 1971).
   It has been suggested that wine may have had some particular significance in the funerary ritual, and many burials containing amphorae have been found in Roman and pre-Roman graves and that the presence of wine may have transcended ostentation or refreshment of the deceased (Salway 1988). At Otford, 36% of the graves contained flagons or bottles, had cups, while 17½% had both (see below).
   After the funeral, relatives underwent a purification rite by fire and water, cleansing ceremonies were held at the deceased's house and a funerary feast was held at the grave in honour of the dead. Later, offerings were made at the grave including a libation to the Manes, and throughout the year funerary meals were partaken as at the deceased's birthday and festivals of the dead (Toynbee 1971).
   As Alcock observes 'the surviving evidence, in fact, suggests that in Britain the toleration which Romans showed towards religious beliefs extended equally to rites relating to the formal disposal of the dead. Celtic religious ideas are seen to continue throughout the Roman period even after the adoption of Christianity, and personal choice was permitted in both religious and burial practices (Alcock 1980). Thus Southern Britain rapidly settled down to Roman rule, and rural dwellers undoubtedly continued to live in much the same way as hitherto, with a nominal respect for Rome's gods and the 'dull divinised emperors', whilst perpetuating their veneration for the old proven gods of their own tradition (Ross 1967).
   The tradition of cremation apparently gained popularity to the point of virtual exclusivity in the first century A.D., and remained predominant in Britain until the 3rd century. The earliest recorded Romano-British inhumation cemetery identified so far was found at Chester, dated to the second half of the 2nd century being closed c 200 A.D. (Collingwood & Richmond 1969).
   A further possible reason for the changes from cremation to inhumation, or vice versa, is cultural rather than religious, as when, within the last 100 years, British funerary practices have changed from 100% inhumation to 50% cremation (Drewitt 1988).
   The polytheism of the Roman Empire developed through absorption of the religious beliefs of the many peoples who were incorporated into it. With the notable exceptions of Druidism, Judaism and Christianity, it appears to have tolerated most of the religions it encountered, and in many instances linked the local gods with the Roman pantheon, the interpretatio Romana, thus, Jupiter, the chief Roman god, was linked with the Greek Zeus, and Juno the chief Roman goddess, was linked with her Greek counterpart Hera.
   In Britain, inscriptions have linked Minerva with the Celtic goddess Sul (Sulis) at the Bath geothermal springs, also with the Brigantian tribal goddess Brigantia, and Mars with Medocuis, apparently an East Anglian British deity, (Somerset Fry 1984), elsewhere the equation appears as Mars Lenus, and Mars Anextiomarus (Frere 1973), and numerous other correlations have been identified.
   In addition to the indigenous Celtic religions, mystical eastern religions were transmitted throughout the empire by military personnel, traders and possibly slaves, which are represented in Britain as in other provinces by temples and altars, statuary and regalia e.g. at Walbrook, London and Carrewburgh, although the rituals of some were eliminated by the early church as parodying the death and resurrection of Christ.
   The varied nature of grave-goods (see below) indicates a conscious attempt to equip the spirit of the deceased for its journey into the underworld, possibly after a period of confinement to the grave, and whereas some objects of considerable intrinsic value were interred or burnt on the pyre to judge from extant remains, the possibility cannot be excluded that the belief was less than wholehearted, at least in some instances.
   Among references to the superstitious beliefs in Roman society in the works of classical authors Alcock cites Heroditus, where Perandius' wife complained from the otherworld that she was cold because her clothes had not been burned with her (History v.27) (Alcock 1980).
   In a late first century grave at Lankhills cemetery, Winchester, the remains of a meal including a young pig and poultry were recovered, and evidence among the interments of some weird and outlandish funerary ritual, including decapitation, suggested to ease entry to the otherworld (Green 1986). It has not been possible to identify any traces of food or drink in the vessels accompanying the urns at Otford, but such could well have been destroyed by the soil conditions, although some instances of vessels having been inverted at deposition, and one (Group 45) with three vessels standing one within the other, suggest that these, at least, were empty. On the other hand, an early second century cremation at Canterbury castle included a samian dish which contained the remains of a small bird (Bennett et al 1982).
   It is clear from both surviving inscriptions and the writings of Roman authors that the religions of the Roman world were abounding in superstition, and that burial practices were intended to ensure that the spirit of the dead person did not return to haunt the living (Alcock 1980), a sentiment expressed by Ovid. The numerous references to the symbolic breaking of grave goods, a practice traced back as far as the La Tene period (Green 1986), indicate that the symbolism continued into Roman times, and various authors (e.g. Jessup 1955 and Alcock 1980) point to the Celtic and Roman belief in 'killing the life spirit' before placing objects in the grave or on the funeral pyre. It is also possible that sometimes it may have been intended to discourage or prevent re-use after ritual dedication (Salway 1988). At Portslade, Sussex, associated with one of the cremation groups in a Roman cemetery, the skull of a sub-adult/hornless sheep was unearthed, considered to be intended either as nourishment for the deceased or as a ritual deposit to facilitate entry to the netherworld by the deceased (Gilkes 1988). Vegetable matter seldom survives in cremation burials but 'walnut or filbert shells' were noted with cremations in the wet conditions within the Bekesbourne shaft (Brent 1859).
   Black draws attention to an apparent development in the scale of provision of eating and drinking vessels, and suggests that 'the generous provision of feasting equipment in Welyn-type burials and the lavish multiplication of pots in first century Romano-British cremations seems to attest a strong belief in the need for sustenance within the grave. Out of this there seems to emerge something like a standard set of vessels : cup/beaker, flask/flagon, and bowl/jar/plate. He goes on to point out that vessels are rarely found after the mid-fourth century, and postulates that the burial of pots and other containers appears to have become a customary rather than a religious duty by then, perhaps earlier (Black 1987). This would tend to reinforce the suggested influence of Christianity by the fourth century on the inhabitants of Roman Britain, with rejection of continued earthly needs associated with the doctrine of resurrection.
   A possible analogy in modern Chinese (Hong Kong) society is that paper models of wealth symbols, e.g. houses, cars, and "Hell bank-notes" are burned on funeral 'pyres' (minus the body) to support the deceased in the afterlife without detriment to the wealth of the living (personal observation). An intriguing possibility arises from the foregoing, that a somewhat similar view was abroad in Roman Otford, suggested by the inclusion of imperfect vessels as grave-goods (see below).
   Recent studies have concluded that there is significance in the alignment of inhumation graves (Kendall 1982) and it is equally possible that there was relevance to the community in the arrangement of ancillary pots and objects placed with urns. At the St. Pancras cemetery at Chichester the excavators noted three instances where the vessels were ‘arranged in a semi-circle with a flagon or dish opposite and the bones scattered between’ (Down & Rule 1971). However, at Otford the plough damage and incomplete investigation of the cemetery diminish the chances of establishing a reliable or even recognisable pattern.
   At Otford there have been no identifiable graves of infants or small children, although mortality of such is likely to have been fairly high during the Roman, as in earlier and later, periods. Numerous burials of infants within villas e.g. Lullingstone (Meates 1979) farms e.g. Sedgebrook, Plaxtol (Bishop 1990), and temples e.g. Springhead (Penn 1967, Jessup 1970), to quote local examples, are generally regarded as foundation deposits, and likely to have been available products of infant mortality rather than human sacrifices (Henig 1984).
   At King's Wood, Sanderstead, a cemetery was discovered just outside the entrance to a small farming community extant in the first and second centuries. It contained five cremation interments of babies and small children and appears to be exclusive to the young, with a separate cemetery for adults further away (Little 1961). An excavation trench at Canterbury castle yielded a late first to early second century cremation group which proved, upon analysis of the bones in the urn, to be the remains of both a 4/5 year-old child and a 'not elderly' adult (Bennett et al 1986). A group at St. Pancras with 8 vessels including 2 flagons contained both an iron knife and a baby's feeding bottle, suggesting the presence of an adult and a child (Down & Rule 1971).
   That not all infants were accorded the same burial rites as adults is indicated by a baby who apparently died at or around birth and was buried in the upper fill of a rubbish pit at Bullock Down, Sussex (Drewett 1988). The reasons for this unceremonious concealment must remain a matter for speculation.
   In a fourth century context at Springhead the excavator interpreted a small flint-built structure as a mausoleum, having a young child inhumated in its chalk floor. Adjacent to this he tentatively identified 'a longish tiled feature (some 7ft. by 2ft.) as a platform for conducting cremations', there being extensive cracking through heat visible and a number of infant burials and cremations in proximity. Other interpretations are possible, however, as there are traces of industrial occupation in the immediate vicinity. A significant general comment on Springhead is that no adult burial has been found in the town, which suggests that rules and the boundaries of the town were still being observed at a late date (Penn 1961).
   A possible significance is suggested for the placing of shoes in relation to the cremation urn, akin to that postulated for orientation of inhumations in cemeteries (Black 1987); at Otford one example (group 34) was noted where a quantity of badly corroded hob-nails was situated to the north of the urn. A logical explanation for the presence of footwear (sandals and boots being noted) is that they were intended to aid the deceased to walk to the underworld. It has been suggested that in some instances only a few nails were placed in tombs as a symbolic gesture. A pair of purple shoes was associated with a rich cremation grave at Springhead (Henig 1984), while at Cirencester a furnace and over 2000 hob-nails were found within a building inside a cemetery conjuring up visions of a 'dedicated' industry (Salway 1988).
   Otford produced a few items which may point to a lessening in belief in, or requirement for, grave-goods even by the mid-second century, although an alternative explanation, that of poverty, is tenable. The creamware flagon (group 72) was undoubtedly a kiln waster, having cracked and distorted during firing to the extent that it could not have held a liquid. It is conceivable that the crack could have been filled e.g. with clay as a temporary expedient, but it remains an extremely sub-standard piece. The St. Pancras cemetery produced 'large numbers of vessels with burials (which) were wasters, and this could be another example of using a "killed" object for funerary purposes' (Down & Rule 1971). A samian plate (Dr 31) from Otford (group 31) had been broken and mended by means of lead rivets, a not unusual repair to samian plates, prior to its deposition. The significance, if any, of the plate in its repaired state is not possible to assess, as it is equally likely to have been the favourite/personal crock of the deceased as to have been placed in the grave for any other reason. Another samian item, a cup (form Dr 33) from nearby (group 70 ) was discovered broken in situ, but on restoration was found to have a piece missing from its rim, and from its position it is likely that this vessel was also imperfect when buried.
   As mentioned above, the supply of libations to the grave both inhumation and cremation was a part of post-funerary ritual, and in many places throughout the Roman empire, including Britain, a pipe was constructed to permit the ingress of food or wine to the grave itself. Lead pipes were sometimes used for this purpose, as at Caerleon (figure lb), while at Chichester a pipe was constructed from tiles - imbrices (Down & Rule 1971).
   The high incidence of flagons with interments has been noted, and a number have had their tops missing. It is known that at Ostia flagons were deliberately left with their necks proud of the ground to receive libations, and a similar suggestion is made for some at Chichester (Down & Rule 1971) although this does not appear to be the case either at Ospringe (Whiting et al 1931) or at Otford, where a high proportion of flagons have retained their upper extremities.
   According to Pliny the Elder (Naturalis historiae xxxv) the majority of mankind employs earthenware receptacles for this purpose i.e. burial, and pottery kitchen jars were the most ideal and cheapest containers for cremated bones and ashes, but other vessels were also used: glass, wood, metal especially lead, and stone (de la Bedoyere 1989). Indications of wooden boxes or caskets are common and were found at Otford, Ospringe and St. Pancras, One such, from Skeleton Green, Braughing, Hertfordshire, is illustrated (figure la). Ospringe and St. Pancras both provided evidence of glass vessels, unlike Otford, but the last-named may have contained at least one lead coffin (see below). Most lead coffins or ash-chests were rectangular, but a cylindrical one containing the remains of a little girl was discovered at York (Fig. 1c).

Figure 1  a) Reconstruction of wooden casket having metal fittings and dating from the mid-second century. Width c30cm.
 From a cremation burial at Braughing, Herts,   b) Pipe burial with a lead canister in a tiled cist from which a lead pipe led to
the ground surface. Caerleon.  c) Lead canister urn with remains of child and inscription. York. (after de la Bedoyere 1989).

   Pottery lamps were provided in numerous graves, presumably to either illumine the deceased during a sojourn in the grave or on the journey to the underworld. Although present in significant numbers at St. Pancras (36) they do not occur at either Ospringe or Otford.
   Jewellery is found in only a small minority of graves, and Otford is no exception, with only three groups providing any traces. Perhaps significantly, two out of these were not furnished with any ancillary vessels.
   It is clear from the writings of Roman authors that the Gauls, who were Celts and closely linked to the British of southern England, had very strong belief in reincarnation, and Julius Caesar commented that the Druids wished to convince men that their souls do not perish, but go from one body to another when they die (de bello Gallico vi), to the extent that they regarded death as unimportant, the Gauls believing that men's souls are immortal and they return to life after a prescribed number of years, with the soul entering another physical body (Diodorus Siculus v 28).
   Black suggests that Caesar's comment on the similarity of the customs of the inhabitants of Cantium to those of Gaul may be taken as a guide to the situation in South-east England in the mid-first century B.C., by which time cremation seems to have been the norm. However, his suggestion, following Whimster (Whimster 1981), that insecurity resulting from Caesar's campaigns may have influenced the change towards cremation (Black 1987), is only one of a number of possible reasons for the practice, not the least plausible being changes in theology of the Celts of Gaul and Br tain. As one authority commented, we actually know less about the religion of the pagan Celts than certain scholars, past and present, would have us believe’ (Ross 1967). This is due partly to the subsequent authority of the Christian church, and partly to the ubiquitous practice among the pagan Celts of instruction by symbols and enigmas, or dark allegories, by ancient songs and maxims, orally delivered and in private, by which they deemed it unlawful to reduce into writing, or to communicate to any but their own order (Meyrick 1848).
   Thus the destruction of the Druidic priesthood caused the loss of their philosophical traditions, which Ross concluded to have been 'little different from the priest/shamans of the entire barbarian and later pagan world concerned with shape-shifting and primitive magic, controlling ritual and propitiating the treacherous gods with sacrifice’ (Ross 1967).
   Pomponius Mela, writing cA.D.43 records one such maxim, a triad, apparently a form of Celtic catechism:
                    "To act bravely in war
                     That souls are immortal
                     And that there is another life after death"
                                                                                 -De Situ Orbis.
A similar triad is preserved in the writings of the third century Greek author Diogenes Laertius:
                      "To worship the Gods
                       To do no evil
                       And to exercise fortitude"     (Meyrick 1848).

   Doubtless many superstitions connected with the natural world are pre-Christian survivals, and some, such as providing a symbolic coin or coins in a burial, are directly descended from Roman practices relating to payment to Charon for the journey by boat to the underworld, and to earlier Greek and Egyptian beliefs. Black, commenting on the low percentage of Romano-British graves containing coins (2-6%) and even lower of cremations (where survival of coins is more doubtful), suggests that the custom appears not to have been adopted by the Celts in Britain, and, where found suggests Roman belief and practice (Black 1987).The lack of any trace of coins at the Otford cemetery, although they occur nearby as field scatter, reinforces the premise that the graves contain the mortal remains of people adhering to local traditions, in line with Ross’s conclusion that 'with the imposition of Roman rule and the creation of the Civil Province of Britain, the south soon settled down under Roman rule and the country people no doubt continued to live in much the same way as before, with a nominal respect for the gods of Rome and the dull divinised emperors', and a continued veneration for the old, proven gods of their own tradition (Ross 1967).
   A Celtic preoccupation with the human head is undeniable, with numerous literary allusions to the miraculous survival of severed heads, e.g. the legend of Bran the Blessed. (G. Jones & T. Jones (trans.) 1974). Votive offerings of heads, both real and carved, have been recovered from sacred sites throughout the Celtic world, and a stone doorway excavated in Roquepertuse, Gaul, was decorated with relief stone-carved heads interspersed with human skulls rebated into the stonework (Delaney 1986), thus demonstrating the intended visibility of such and thereby adding weight to the observations made 'in Graeco-Roman literature (Green 1986).
   There are ample indications that burial grounds were intended to be continuing reminders to the living of those who had gone before, hence cemeteries must have been demarcated in some enduring manner, with either boundary ditch, posts, hedge, fence or wall. Walled cemeteries and funeral monuments within walled enclosures are restricted almost entirely to the South-east, with eight of the fourteen identified being in Kent (Jessup 1970), and the Otford cemetery was provided with one such monument, a mausoleum, although limitations of excavation have not to date permitted a search for a boundary. Nonetheless, the proximity to a contemporary trackway strongly points to the need for some form of delimitation, perhaps a thorn hedge, if only to keep out wandering cattle.
   Personal gravestones or markers may be expected in view of the fairly regular spacing of the graves in most Roman burial grounds, strongly suggesting that individual locations were marked for posterity; in no instance at Otford were graves superimposed as would be likely if markers were absent. Indeed, the presence of a 'stake-hole' within one of the urns (group 2) might represent such a grave-marker, but could equally well date from any subsequent date. However, a small post-hole close to an urn at St. Pancras (group 146) contained a mass of charcoal which 'may have been the charred base of a marker post’ (Down & Rule 1971) and the tentative identification of a simple late-first century cremation grave at Eastwood, Fawkham, as being marked by a stake (de la Bedoyere 1989), do reinforce the marker theory. Also, one of the two grave groups at Canterbury castle had 'just to the north, four tile fragments stacked one upon the other, possibly the base of a grave marker', these graves being probable survivors from a larger cremation cemetery destroyed during the building works of the 1950s (Bennett et al 1982).
   In addition to cemeteries, Britain has the remnants of a number of upstanding Roman funerary monuments, the dating of many of which is obscure. The most numerous class is the round barrow of which some one hundred are known, and which has an antecedence stretching back to the Neolithic period. It persisted through the Iron Age, notably in the great round Belgic barrows at Lexden near Camulodunum, and occurs in Roman Britain primarily in the South East, where circular conical barrows are found singly and in groups, notably at Bartlow in Essex, where five barrows contained wooden chests enclosing cremations and one covered a tile-built burial cist (Toynbee 1971).
   At Holborough near Snodland, a barrow, known locally as Holborough Knob, containing a particularly richly furnished cremation was excavated in 1954 in advance of chalk quarrying. The bones were placed in a wooden coffin placed in a grave over which a puddled chalk dome was erected initially, then enclosed in the barrow-mound. The funeral pyre was elsewhere, from whence grave goods: potsherds, fused glass, burnt bones including those of a (?) sacrificed cockerel, a memorial coin to Antoninus Pius (died A.D.161) depicting a funerary pyre, and the frame of an iron and bronze folding stool were brought. The funeral ceremony is deemed to have been conducted from a temporary shelter and included a libation of resinated wine. An unusual feature is the insertion of a secondary, infant-female, inhumation with the coffin decorated with Dyonysiac figures. Stylistically, this dates from the early third century (Jessup 1955, 1970).
    A round barrow at Plaxtol, described as a large hillock with a covering shaw, some 28 ft. in diameter, containing a central inhumation and a number of secondary Roman cremation groups around the circumference, was grubbed out in the mid-nineteenth century. At the same time it was concluded that traces of a foundation nearby represented a wall delineating the cemetery (Luard 1859).
    A development of the 'earthen barrow' after the Italian style, is the enclosure of the barrow within a low masonry drum. Of this type is considered to be a round tomb at Keston with a 3 ft. wide flint wall without an entrance, having six radiate external buttresses (Fig 2c), which, besides resisting the outward thrust of the barrow, may have supported low ornamental pillars (Toynbee 1971). Alongside this was a rectangular tomb chamber containing a stone coffin (Detsicas 1983), and also a cremation inside a rectangular lead casket within a tile tomb (Jessup 1970).

Figure 2 a) Otford mausoleum, b) Lullingstone complex (after Meates 1979). c) Keston tumulus & tomb. 
d) Pulborough mausoleum ( after Collingwood & Richmond 1969). e) Lancing Down temple (after Bedwin 1981).
f) Mausoleum Welwyn (after Rook et al 1984). g) Langley walled cemetery. h) Springhead walled cemetery 
(after Detsicas 1983) i) Titsey temple (after Graham 1936).

    Introduction      To Frog Farm Cemetery contents page      Frog Farm Site