Where Different Waters Met Aspects on the Apollo Grannus vase and its position near Sagån in the province of Västmanland Lotta Fernstål

At some time during the Late Roman Iron Age, an originally provincialRoman vessel called the Apollo Grannus vase was used as a burial urn in a stone-setting at Fycklinge in the province ofVästmanland, on the bank of Sagån by the small rapids where this stream met the sea. In an attempt to understand why this particular vessel was used in this way at this particular place, the author discusses the characteristics of the area, ofcertain vessels and their associations with rebirth/regeneration, the provincial-Roman god Apollo Grannus (associated with fertility and water), the process of creolization, and cremation as part of a process of transformation.

The vessel is 45 centi metres high, made of bronze but origi nally gi lded, and ornamented wi th silver and possibly also copper inlays forming a borderjust below the &i m, as well as flourished motives on bothsides oftheinscription.Photo from Stenberger l956.44.
of Sagån in this particular area (ATA a, c, d) (Fig. 2).The exact spot where the vessel was found is not known, but in the registration of ancient monuments by the Nati- onal Heritage Board the approx- imate area was marked with the number 176 (RAÄ 176), Björksta parish (all the RAÄ-numbers lacking information on parish in this text are from this same parish).If we once again tum to the earliest sources concerning the Apollo Grannus vase, this approximate area can be narrowed down to the northwest corner of a modern garage and, near this corner, the remains of a foundation of a cottage which once belonged to Lundström who found the vase, possibly while he was building this same cottage (ATA d).After the find was made, Lundström made a memorial stone out of half a millstone (RAÄ 177) to commemorate the event.On this millstone, which today is not situated exactly where the find was made but in a flower-bed southeast of the house belonging to the garage, Lundström carved the date of the find and a quotation of the inscription of the Apollo Grannus vase, as well as his signature with a flourish (Fig. 3).
In the area where the Apollo Grannus vase was found the landscape opens up to the east, and the small rapids of Sagån, today just north of the modern bridge, catch the visitor's eye.After the rapids, Sagån continues to flow at a slow and steady pace towards its outflow in Lake Mälaren, which was part of the sea during the Late Roman Iron Age, the period in focus in this paper.During that time, however, the fresh water of Sagån met the salt (or brackish) water of the sea already in the rapids by the modern bridge below the remains of Lundström's cottage; the part of the stream which today forms the last part of Sagån on its way to Mälaren, was then namely a narrow arm of the sea.The purpose of this paper is to try to cast light on why this specific vessel was used as a burial urn and placed in a grave at this particular location.In order to understand this, ideas Current Swedish Archaeology, Vol.l/, 2003 Fig. 2. A map showi ng Sagån and the area where the Apolkr Grannus vase was found.The flow ofSagå n today is marked wi th light grey.Note, however, thai the northern part of Sagån on this map ought to have had a simi lar flow duri ng the Late Roman Iron Age.The darker part surroundi ng the southern part of Sagå n shows the area which today fall» below l0 m as l. and which during the Late Roman fron Age was a narrow arm of'the sea.A cross and the RAA number l 76 mark the approxi mate place where the Apollo Grannus vase was found.The road» are modern.The map wa» drawn by the author and based on the economical map and the Register of Anci ent Monuments.
about Apollo Grannus, vessels with associations, the process of creolization, a possible meaning of cremation as part of a process of transformation, and the character of the place and area where the Apollo Grannus vase was found, will be discussed.1 will start, however, by giving a background to the dating of the Apollo Grannus vase, the circumstances of the find and, briefly, the tendencies of the earlier scholarly research concerning this vessel.A few remarks on the origin of the Apollo Grannus vase, and how the vessel may have arrived in Västmanland, will also be given.

A BACKGROUND
The entire Latin inscription on the 45-centimetre-high Apollo Grannus vase reads as follows: APOLLINI GRANNO DONVM AMMILIVS CONSTANS PRAEE TEMPLI IPS IVS VSLLM Interpretations of this inscription differ in details (e.g.Hallenberg 1819:9;Elvius 1930:44 with references; ATA b), but summarily it tells us about Ammilius Constans, who was the principal of a temple dedicated to Apollo Grannus and who fulfilled his vow to the same god with a gift.
The Apollo Grannus vase is, in Scandinavian terms, usually dated to the Early Roman Iron Age and period B (1-150 A. D. ) (Elvius 1930:44; Eggers 1951:165;  Stenberger 1956:47; Lund Hansen 1987:90, 451, 461).The dating of the grave in which it was found is, however, seldom discussed.When the vessel was used as a burial urn, it was already somewhat broken.A lid, two carrying handles and their attachments were missing, and the foot was loose.The silver inlays and the gilding are in poor condition, and the vessel has also become partly damaged by verdigris, which, however, was not necessarily the case when it was used as an urn.Because the vessel was worn, the use of it as a burial urn can be assumed to be of a later date than the manufacturing of it and the use of it in its original Continental context, which, as was mentioned above, can be placed in the Early Roman Iron Age.The utilisation of the vessel as a burial urn may then have occurred in the Late Roman Iron Age.A similar use of "imported", often worn, vessels during this period is known from other places, for example western Norway.
In this area, discussed by Tove Hjerungdal, Vestland cauldrons were used as burial urns (1999).
Fig. 3.A photo and an approxi mate outli ne drawi ng o f the memorial stone whi ch was made to commemorate the find ofthe Apollo Grannus vase.Lundström, who found the vessel i n l 8l 8, made the memorial stone out of haifa mi llstone, and on i t he carved a quotati on from the i nscripti on ofthe vessel, as well as the date of thefindandhis signature.Today the memorial stoneis situatedin a filowet-bedclose to where the findwas made.Photo by the author; drawi ng by the author after a sketch i n A TA c.
Current Swedish Archaeotogs~, Vot.Ii, 2003  It should be mentioned that the find circumstances of the Apollo Grannus vase are not entirely clear.For instance the vessel is sometimes said to have been found in a cairn/stone-setting (Stenberger 1964:389) and sometimes in a mound (Willers 1901:119;Paulys Real-Encyclopädie der... 1912:1826; Montelius 1919:185;  Andersson 2001:225).Sometimes the contents of the grave are listed as fragments of bones and pieces of melted glass (e.g.Hallenberg 1819:80),and at other times there is also mention of a knife, bronze fittings and a garning-piece of bone (e.g.Andersson 2001:225).The different information about the grave and its contents is probably due to the varying statements in the documents of ATA, as well as to a probable confusion concerning objects with the registration number SHM 425 at the Museum of National Antiquities.In short, the knife, bronze fittings and the possible garning-piece seem to be later additions to this registration number, which originally consisted of the Apollo Grannus vase and the pieces of melted glass.What I have done to try to sort out the original find circumstances of the Apollo Grannus vase, is to study the writings of the find chronologically, both the printed material and the archived documents from ATA.I have given the greatest credence to the earliest writings, and the main results of this study were presented in the introduction of this paper.
Leaving the ATA documents about the Apollo Grannus vase aside, this vessel has not often been the focus of scholarly attention.The exceptions are the works of Jonas Hallenberg (1819) and Sven Elvius (1930).Otherwise, since the vessel is one of only two vessels in Scandinavia with Latin inscriptions, it is usually primarily mentioned as a quite special or spectacular object (cf.Montelius  1919:185; Stenberger 1956:46-47, 52-53, 1964:388-389; Ekholm 1961:55-56;  Andersson 2001:221, 224), which is also correct in the sense that it is a unique type of vessel in Scandinavia (Andersson 2001:221, 224).The other vessel with a Latin inscription found in Scandinavia is an Ostland cauldron from Farmen in Hedmark, eastern Norway, a vessel which may have been used in a similar way as the Apollo Grannus vase since it is reported to have contained cremated bones.
The vessel type and the Latin inscription of the Apollo Grannus vase may make it a unique object, but the phenomenon of using a vessel from a place "far away" as a burial urn, was not unique in Scandinavia during this time (cf.Ekholm  1961:96;Hjerungdal 1999; Holand 1999, 2001).The questions of how and from where the Apollo Grannus vase came to Västmanland have also concerned earlier scholars interested in the find.These questions will be considered below.As is usual in connection with Roman and Continental objects found in Scandinavia, the grave with the Apollo Grannus vase has been suggested to be an expression of power and wealth (Stenberger 1956:52).The person(s) buried in the grave may have been both powerful and wealthy, but this does not, however, explain why this specific vessel was used at this particular place.
Loua Fems(å I FROM A PLACE "FAR AWAY" As was mentioned above, how the Apollo Grannus vase ended up in Västmanland and from where it originated are questions which have concerned earlier scholars. Suggestions of how the vessel ended up in Västmanland mostly centre around a supposed plundering of the temple to which the vessel was originally donated (cf.Montelius 1919:186;Elvius 1930:44; Stenberger 1956:47, 1964:388; Anders- son in  Andersson & Herschend 1997:33), but also trade (Willers 1901:120).According to Elvius, it was a farmer who had participated in the plundering of the temple, who was buried in the vessel (1930:46; cf.Montelius 1919:186-187;  Stenberger 1956:47).
Regarding the origin of the vessel there are several proposals, which, however, all concern provincial-Roman areas in Europe.The suggestions are mainly based on Continental inscriptions mentioning the god Apollo Grannus, and on where the healing springs which Apollo Grannus is connected with were situated.The inscriptions originate from what used to be the Roman provinces of Raetia, Noricum and Britannia (today the approximate areas of Austria, Switzerland, southern Germany and England), but also from Scotland and the city of Rome (Paulys Real-Encyclopädie der. .. 1895:46; Hallenberg 1819; Montelius 1919:185; Elvius 1930:44).Apart from the inscriptions, there is a relief from Altbachtal at Trier in western Germany which is said to represent Apollo Grannus leaning against his tripod and holding a vessel, from which flows healing water.Further, at the healing springs at Aachen, also in western Germany, Apollo Grannus was wor- shiped as the main god.Apollo Grannus was, however, also worshiped at springs in other parts of the Eifel area between Aachen and Trier (Elvius 1930:45).
Several of the areas with inscriptions mentioning, and springs connected with, Apollo Grannus coincide with areas in which Celtic people are considered to have lived, and Apollo Grannus has also been suggested to be of Celtic or Celtic- Roman origin (e.g.Montelius 1919:185-186;Elvius 1930:45; Schuck 1943:306;  Stenberger 1956:47, 1964:389).During the last years, however, the concept of Celts has been discussed critically (e.g.Chapman 1992; Wells 1999:264, 2001), and it does not go without saying that the Celts can be regarded as a homogeneous ethnic group whose exact areas can be located.Perhaps it is enough to establish that the Apollo Grannus vase probably originates from the Roman Empire, presumably from one of the provinces in Europe since these are the areas where most of the other Apollo Grannus inscriptions have been found.But whether the vessel may be called Celtic-Roman is not as interesting as the fact that it came to Västmanland from a place "far away".

A PROCESS OF CREOLIZATION
During the Roman Iron Age the so-called futhark and the runic script were invented.
Some runic inscriptions from the second half of this period seem to be the personal names of objects, for example a couple of spearhead inscriptions which read tilarids, Goal-pursuer (Antonsen 1975:74) or The Assailant (Snadal 1994:9), and raunijaz, The Tester or The Prober (Antonsen 1975:29).These inscriptions reveal that, during this time, certain objects were believed to have a kind of "personality" and their own, individual character.One way to understand why a particular object was used in a certain way, in this case why the Apollo Grannus vase was used as a burial urn in a grave at a location where Sagån met the sea, is to try to find out what kind of ideas and associations that may have existed about that specific object.These ideas and associations may be difficult to reach, but with respect to the Apollo Grannus vase, a worthwhile point of departure is to discuss provincial-Roman ideas about Apollo Grannus.In light of the Apollo Grannus inscription, these ideas may have been connected with the vessel and also "followed" the vessel on its way to the place where it was finally used as a burial urn.
I must, however, point out that by this I do not mean that a transmission of provincial-Roman mythology had occurred, that people in Scandinavia believed in a god called Apollo Grannus.Neither do I mean that provincial-Roman ideas about this vessel had been transmitted to Scandinavia; when the vessel was in its original context, there probably did not even exist ideas connected with the inscription that were similar to the ones I believe may have existed in Scandinavia.
Ideas about the vessel connected with Apollo Grannus were rather inventions by people who got the vessel in their hands at some point after it had been taken from its original setting in a provincial-Roman temple.They were people who had knowledge of provincial-Roman beliefs but who created something new out of this knowledge.This is an example of a process called creolization, a concept which may need some explanation since it originally was a linguistic term used for processes leading to the creolized languages.
Creolized languages are, simply put, "mixed" languages with words and grammar that diverge from "original" languages.However, the concept of creoliza- tion is also used in a wider sense for the meetings of cultures and traditions which result in the creation of something new (Friedman 1995:208).A quotation that well represents the meaning of the concept is: ". ..all cultures are involved in one another; none is single and pure, all are hybrid, heterogeneous, extraordinarily differentiated, and unmonolithic" (Said 1993:xxix).During the later years, theories of processes of creolization have been important parts of post-colonial studies concerning prehistoric conditions in Roman provinces of western Europe (cf.Webster 1997, 2001; Woolf 1997, 1998).Important results of such studies show that, in these areas, new traditions were developed in combination with both local and foreign influences.The Roman Empire came to comprise a large number of diverse areas with individuals who met and were influenced by each other in different ways.This should, however, also have been valid for people from outside the borders of the Roman Empire.People from both outside the borders and from inside the Empire travelled both ways and met and influenced each other.One such case of creolization may then have been, as mentioned earlier, that Scandinavian ideas about the Apollo Grannus vase were influenced by provincial-Roman Current Styedisit Areitaeology, Vol.ll, 2003 Lotto Fernetå I ideas about Apollo Grannus, and therefore the vessel came to be used in a certain way at a certain place.I will return to this further below.

APOLLO GRANNUS, FERTILITY AND WATER
As mentioned above, the Apollo Grannus vase has earlier been suggested to have been brought to Scandinavia and Västmanland through trade or directly from the plundering of the temple where it originally was dedicated to the god Apollo Grannus.An alternative scenario is that the vessel reached Scandinavia and Västmanland as a gift or as part of a dowry (cf.Holand 2001:83-84).In connection with this, another possible scenario is that during gift exchanges (and perhaps also during trading situations) stories and ideas about the exchanged objects were told which made at least some of the objects even more special and precious.
Further, if the vessel arrived as part of a dowry, a situation where stories and ideas about the vessel were told to persons in its new surroundings, is conceivable.As noted above, these stories and ideas may have followed the objects and been important parts of them, but they can also have been somewhat altered to suit the new owners, or been interpreted in different ways in different areas.For this reason, it is not possible to use the provincial-Roman ideas about Apollo Grannus as a point of departure and single out one idea about the Apollo Grannus vase in Scandinavia.But by discussing such ideas, it may be possible to reach some understanding of the character of the Scandinavian ideas and associations that may have existed behind this vessel.What then were the provincial-Roman ideas about Apollo Grannus?Several characteristics of Apollo Grannus have been suggested, just as there seem to have been several different aspects of Apollo.Apollo was originally a "Greek" (another concept that should be discussed critically regarding prehistoric times) god, among other things connected with the oracle shrine in Delphi, poetry, music, prosperous agricultural production, the creation and extinction of plague, healing/curing and purification, athletic training and military violence (Bruit Zaidman k Schmitt   Pantel 1995:193-196;The Oxford Classical. . . 1996:122; MacKillop 1998:230;  The Oxford Companion. . .1998:51).Even in the Roman Empire, where Apollo was also worshiped, several features were probably linked to the god, although his main feature seems to have been as Apollo Medicus, that is as a curing or healing god (The Oxford Classical. .. 1996:123;also cf.Hallenberg 1819:26, 51).
Hallenberg has different suggestions for the origin of the byname Grannus, for instance that it should be sought among "Geats and Teutons" (1819:23),that it can be of "Celtic" origin, (1819:25)or that it may be derived from the "Slavonic Gran, border or boundary mark" (1819:35-36,my translation).Moreover, Hallen- berg proposes that there can also be a connection between the byname Grannus and the Latin words granum and granarium, corn and granary (1819:45), or, if Grannus can be derived from a Geatish or Teutonic word, the word granne, which was used for hair like rays (1819:56) (in connection with the sun).The last- mentioned can be compared to information in Paulys Real-Encyclopadie der. . ., Current Sn~edieh Arehaeology, Vol.I I, 2003 where Grannus is connected with the Gaelic word grean and the "new-Irish" granni, which mean hair, beard (1912:1826).Further Hallenberg maintains that Apollo Grannus can be considered as "twice the name of the Sun", as Sol Mithra or Sol Serapis (1819:20).Oscar Montelius also associated the byname Grannus with the sun, and meant that Grannus could be the Latin form of a Celtic word signifying sun (1919:185),possibly indicating the Irish word grian.Miranda J. Green (1992:32) and James MacKillop (1998:230) believe, however, that the name Grannus can not be linked philologically with that word, but that the god nevertheless possessed a solar aspect.The same view is held by Miklos Szabo (1971:66)and Lesley Adkins k, Roy A. Adkins (1996:14).
Both Green (1992:32) and MacKillop (1998:230) suggest that the name Grannus corresponded to the ancient name for Grand in the Vogses in eastern France, where there is assumed to have been a form of cult centre like the one at the healing springs at Aachen, mentioned above.Elvius, instead, connected the name Grannus with the word granum, but in the meaning of "granule which originates from spring water" (1930:45, my translation) rather than grain, and he referred to the springs at Aachen whose Latin name, according to him, was Aquisgranum and where Apollo Grannus was the main god.At springs the sediment is often of a granular structure, which, according to Elvius, should have been the case also at the iron, sulphur and salt springs at Aachen (1930:45).But according to the ATA-document Apollo Grannus-vasen från. . .(ATA b), the name Aquisgranum was a medieval idiomatically incorrect form of Aquae Granni, which was the original Latin name of Aachen.That Aquae Granni was the ancient name of Aachen is the opinion of Green as well, who suggests that the name means "the waters of Grannus" (1992:32).Bernhard Maier also believes that Aquae Granni means "the waters of Grannus" and that this name is the Latin name of Aachen, but he further says that this name was not attested until 765 A. D.
(1997:136).Other information concerning these names indicates that the classical name of Aachen was Aquisgranum, but that the springs were called Aquae Grani (Nationalencyklopedin 1989:1-2).The intention here is not to make an additional attempt to solve the question of the original name of Aachen or the healing springs there, but it is worth mentioning that other scholars have also associated Apollo Grannus with both Aachen and healing springs at other places.Szabo (1971:66), Green (992:32), Adkins k, Adkins (1996:14) and MacKillop (1998:230) mean, for example, that Apollo Grannus was a healing god and associated with what are also sometimes called therapeutic or medicinal springs (also cf.Paulys Real- Encyclopädie der. . .1895:46 and Paulys Real-Encyclopadie der... 1912:1827).
In the view of the above discussion of the name Apollo Grannus, several aspects of the god seem to have existed, just as they did for Apollo.Whether Apollo Grannus was some kind of hybrid between Apollo and another god called Grannus, with their respective characteristics, or if Apollo Grannus was a variant of Apollo, is not easy to determine, and perhaps not even necessary.The result might be almost the same either way; that Apollo Grannus can be seen as a variant Cn&rent Sieedivl& A&ct&aeotog&:, Vot.ll, Ztl03 of Apollo with some common and some diverging aspects.What should be mentioned in particular as a common trait is the healing/curing aspect, which at least in the case of Apollo Grannus seem to have been connected with water and springs.Noteworthy is also that the water in such springs is characterized, among other things, as being hot and bubbling, which is clear for example from the Internet homepages of spa resorts and clinics in the Aachen area.The connection between Apollo Grannus and water will be discussed in more detail below.
Something else which has possibly characterized Apollo Grannus is a connection with the sun and corn/granary and, in a wider sense, with fertility.That the last- mentioned interpretations of the characteristics of Apollo Grannus are based on words from different languages is not necessarily a problem, partly because the meaning of the name may have varied somewhat in different areas, and partly and perhaps primarily because this name can also be seen as the result of a process of creolization.
The complete name Apollo Grannus may then have been a fusion of different aspects of the god Apollo and of the byname Grannus, created of words with similar stems deriving from different languages.These words have diverse meanings, but some of them can be linked together and their common character summarized in the concept offertility.Further, as mentioned above, Apollo Grannus was also associated with water.I will return shortly to these aspects of Apollo Grannus in relation to the Apollo Grannus vase.First, however, "general" ideas and associations behind certain vessels used as burial urns in Scandinavia will be discussed.As stated earlier, the use of vessels from "other places" as burial urns in Scandinavia was not a phenomenon unique for the Apollo Grannus vase.
Perhaps even more important than the fact that the vessels came from "far away", however, is their association with beverages.

VESSELS, FERTILITY AND REBIRTH/REGENERATION
In classical terms (valid for both the Greek area and the Roman Empire) the Apollo Grannus vase would be defined as a ctater (SHM 425 in Tillväxten; Hallenberg 1819:62), which is a large vase with a wide mouth.In a crater water and wine were mixed and then served in smaller cups, both at banquets and sacrifices as well as libations to gods (cf.Mollett 1996:94; Der neue Pauly. . .1999:807).In terms applicable to the northern parts of Europe, the Apollo Grannus vase can be compared to the so-called Hemmoor buckets (e.g.Ekholm 1961:55-56, 1974a:97-98, 1974b:223 where Ekholm maintains that the Apollo Grannus vase is a "transitional type" between classical vessels with the attachments of the carrying handles soldered on and the actual Hemmoor buckets (also cf.Eggers 1951:165))(Fig.4).Hemmoor buckets may also have been used to mix drinks in and serving them from.
Like other "imported" vessels used as burial urns during the period in focus (Hjtårungdal 1999:81, 89; Holand 2001:77), the Apollo Grannus vase is partly broken.As Hjerungdal writes, the use of worn vessels as urns may have its basis Current Swedish Archaeology, VoL ll, 2003 Fig. 4. The Apollo Grannus vase can be compared to the Hemmoor buckets, whose main types are outlined here.
A ccordi ng to Gunnar Ekholm, the Apollo Grannus vase as well as two other vessels found in Scandinavia was act«ally a "transitional type" between classical vessels and Hemmoor buckets (l974b: 222-223).The first vessel at the top left in a belief that it was important to use old vessels which earlier generations had owned (1999:89).Ingegerd Holand also discusses the use of "imported" vessels as burial urns during the Iron Age in western Scandinavia.She points out that, since there are only a few vessels which are sooty, cooking vessels were probably not used as urns to a great extent; rather, the urns were mainly vessels that had once been used for beverages, and they did not leave visible traces like cooking did (1999:156,2001:78).This was possibly also the case with the Apollo Grannus vase, which, as mentioned above, originally was a crater for the purpose of mixing drinks and serving drinks from, and which also can be compared to the Hemmoor buckets.
ln light of so-called Celtic (Continental might be a more suitable word) ideas and associations behind vessels, which Holand believes may have been valid in Scandinavia as well, and also the ideas about vessels in Norse Iron Age mythology, Holand (1999, 2001) presents an interpretation of both serving and drinking vessels.Holand reaches her interpretations of vessels through, among other things, discussions about cauldrons as vessels of plenty (cf. the symbolism of Greek, Roman and "Celtic" horns of plenty, cornucopiae (Fig. 5)), which were never supposed to run out of contents at gatherings.In her opinion, during the first millennium vessels had a meaning connected with immortality and fertility, the latter in a wide sense that included prosperity, transformation, rebirth/regeneration, and the continuance of the family.This can be compared to Hjerungdal's theory, mentioned above, about using worn vessels as urns because of the importance attached to vessels that earlier generations had owned (1999:89).Holand also maintains that the reason for using vessels as burial urns may have been their fertility symbolism (2001:95).In connection with that, it can also be mentioned that the link between death and fertility is not unusual; this has been discussed, for example, in Death ctnd the Regeneration of Li fe (Bloch A Parry (Eds) 1994), which consists of papers dealing with anthropological examples of this theme.Fig. 5.A Roman altar-relief from the first century depi cti ng the deified emperor A ugustus sut.roundedbycornucopiaewhich "proclaimhim founder of the general prosperity" (Zanker l 995:308).
What was said above about the symb- olism of vessels ought to apply also to the Apollo Grannus vase and how it was used as a burial urn in Fycklinge.But with regard to this vase, stories, ideas and associations which were connected with the content of its Latin inscription did probably exist in addition to the "general" ideas and associations.Further, as discussed earlier, these were also about fertility in addition to water.It is possible that the ideas concerning fertility and rebirth/regeneration were related to the continuance of the family, as Holand ( 2001) suggests with respect to the meaning of using certain vessels as urns.However, another likely form of rebirth will be discussed below after the phenomenon of cremation has been touched upon.

CREMATION AS PART OF A PROCESS OF TRANSFORMATION
The cremation of a deceased person involves letting the body dissolve into a large amount of fragments.During the Early Iron Age the cremated remains were usually crushed into even smaller fragments, and burials from this period contain a variable amount of bones and often not the amount one would expect if all the remains of a cremated person had been buried in one grave (Bennett 1987:84,  112; Ericsson Sc Runcis 1995:37).There are also constructions which are very similar to those we call graves, but which do not contain remains of humans at all (cf.Ericsson & Runcis 1995).When a construction is judged as a grave, a modern meaning is applied to an ancient material.As Anders Kaliff has pointed out, the interpretation of a construction as a grave is actually quite an advanced inter- pretation, since the stone-settings and other constructions we judge as graves may not have had the function we expect them to have (1997a:2, 1997b:68-69,  118).Thus it is important to keep in mind that the prehistoric ideas behind the constructions we judge as graves might have differed from our own ideas.However, although it might be desired, the word "grave" is in these contexts difficult to exchange for another word, which might be more neutral or filled with other associations.In this paper I will use the word "grave" in spite of the dilemma just mentioned, but hopefully with an awareness of the problems it can cause and with an openness to other interpretations than the ones that stem from modern associations of the concept.
Current Swedish Archaeology, Vot./l, 2003 The fragments of bones found in the Apollo Grannus vase were not saved, and therefore we do not know how small or large the amount of bones was.But since the amount clearly must have fallen within the quantitative range for bones in graves from this period, the grave with the Apollo Grannus vase can also be discussed in relation to this phenomenon.Neither do we know whether the cremated bones in the Apollo Grannus vase consisted of the remains of one or more humans, nor if they included remains of animals.Whether the bones derived from one or more persons may seem important, but in this case and with reference to the comments below about the annihilation of the individual, I do not believe that such a knowledge would have great bearing on the following discussion.
Below I will also return to the possibility that there were remains of animals among the cremated bones in the Apollo Grannus vase.
The cremation and the dissolving of the deceased into fragments may have been part of a process of transformation, the purpose of which was to bring the deceased from one stage of life to another in the "post-physical" life.To cremate and dissolve the deceased into a large amount of fragments, then to possibly further this process by crushing the already fragmented bones, and finally to place perhaps only parts of the body in the grave all of this shows a range of different ways to treat the dead body.And these treatments give the impression of a wish for the definite destruction of the body.It can be noted, however, that the cremation not only led to the definite destruction, but also to a stage in which what was buried was also definitely preserved (cf.Kaliff 1997b:80).Thus, to cremate a deceased person can be seen as a phenomenon which was definitive in different respects; the body was in a definite way destroyed, while at the same time different amounts of the remaining fragments of this destruction were also definitely preserved through the burial.
To dissolve the deceased person into a large amount of fragments can be seen as a way to "deidentify" and "degenderise" the individual (Hjerungdal 1999:86).This enabled the creation of a new identity for the deceased, perhaps an identity as an ancestor, which Hjerungdal (1999), in accordance with Seren Diinhoff  (1997:115),views partly as a mediating link between humans and the world of gods, the powers of nature and other spheres which the world may have been considered to consist of.Alf Ericsson k, Janis Runcis maintain that the cremated and thereby disintegrated person, instead of becoming an individual ancestor, came to be absorbed in a collective of "indefinite" ancestors (Ericsson A, Runcis 1995:37), With this view, empty graves can be understood since they constituted the lower limit of the quantitative range of bones in graves, and they show that an individual was not only transformed but also effaced (Ericsson & Runcis 1995:38).Perhaps both these alternatives (to become an "individual" or a "collective" ancestor) existed, and perhaps there also were other alternatives for what one could become or what stage one could reach after the physical death and the process of transformation.
Curreni Sieedish Archaeolog&:, Voh ll, 2003  In the above, the Apollo Grannus vase and other vessels as well as their possible fertility and rebirth/regeneration symbolism were discussed.It was noted that Holand (2001) suggested that the use of vessels with this symbolism as burial urn, was related to the continuance of the family.Another possible form of rebirth may have been the deceased person's "birth" into a new stage of life after the physical death and the process of transformation.To place (some of) the remains in a vessel which on different levels was associated with aspects of fertility and rebirth/regeneration, could then have been an additional part of the process of transformation and the creation of a new stage of life for the deceased person.
In relation to the "deidentifying" and "degenderising", detailed interpretations like the one mentioned earlier about the cremated person buried in the Apollo Grannus vase being the farmer who had plundered the temple that once housed the vase (Elvius 1930:46), are less important to come up with since they involve the search for a specific person of a specific gender.Neither is the knowledge of whether there originally were remains of one or more persons in the Apollo Grannus vase very important in relation to this.The person(s) could in any case have been absorbed into a collective of "indefinite" ancestors, become individual ancestors, or something else.In connection with this, the possible remains of animals in the Apollo Grannus vase can also be touched upon.One interpretation of such remains is that they are remnants of food for the dead person or of sacrificial meals in honour of the deceased.Erika Räf, on the other hand, has interpreted such remains as cremated animal "helpers" on the journey to another part of the world (2001a:36-37, 2001b:20-21).Possible fragments of cremated animals are in this view, then, also the remains of a part of the process of transformation to bring the deceased into another stage of life, something which may be compared to a kind of journey.

THE BORDERLAND AS A POINT OF DEPARTURE
The Apollo Grannus vase with the remains of a cremated person was placed in a stone-setting close by and with a view of the place where the rapids in Sagån met the sea.This act can be viewed as the last part of the process of transformation, which finally let the deceased person into a "post-physical-death-stage".
The chosen place by the outflow of Sagån into the sea may have been regarded as liminal or as a kind of borderland, because of the meeting of the fresh and the salt waters.Hans Bolin (1999) has reflected on different types of water and their significance for the placing of, among other things, cairns, rock carvings and rock paintings in northern Sweden.According to Bolin, important waterways in different societies and periods of time were given a mythical meaning, for example in order to accentuate the inhabitants' relation to and affiliation with an area (1999:181).Regardless whether Sagån was seen as mythical or not, it had meandered about 60 kilometres on its way to the sea, passing through lands well known to the people of the region.When Sagån met the salt water in Fycklinge, however, the stream met a water whose closest surroundings probably were as Current Styedish Arehaeology, Vol.II, 2003  well known as the areas along Sagån, but which gradually led to the less known and foreign, and to the rest of the world.Various ideas and associations possibly existed about the different waters.Of the fresh water, ideas about the "usual", familiar and well known may have existed, while salt water, because of the possibilities it gave people to travel to almost any part of the world, might have signified the opposite.Also the rapids by the modern bridge in Fycklinge, and the immediate area around these, indicate that this place may have been regarded as liminal.The area is almost empty of ancient remains from the Late Roman Iron Age.There is, however, a cemetery (RAÄ 39) approximately 100 metres northwest of the place where the Apollo Grannus vase was found; it consists of about ten circular stone- settings, although these are possibly of a later date.There is also a stone boulder with about 15 cupmarks (RAÄ 334) approximately 50 metres southwest of the find-spot of the vessel.Furthermore, a Viking Age spearhead (RAÄ 183) was found close to the cemetery.A couple more graves may have existed in the area, to judge from fragmentary information in documents from ATA, but these graves and their age are uncertain.
As mentioned above, today in the area where the Apollo Grannus vase was found there is a modern bridge across Sagån which leads between Fycklinge and Bredsdal.At my first encounter with this place, for different reasons I thought that this 19'"-century bridge might have been constructed at the same location as an old bridge or crossing over Sagån.
There are, however, names of places (Broby, Hanvad, Skarsvad)   indicating crossings of different kinds (bridges and fords) over Sagån within a few kilometres both to the north and south of Fycklinge (Fig. 6).There are also remains of stone constructions registered as Current Swedish Archaeology, Vol.ll, 2003  fords in the area (RAÄ 178 in Björksta parish, Västmanland, and RAÄ 236 in Breds parish, Uppland).Since these names and constructions indicate several possible places in the area where Sagån could be crossed, and since no names or information about ancient remains indicate a bridge or ford at Fycklinge, I abandoned the idea of this place being an old crossing.
There may be a practical explanation for why this place was not an ancient crossing, but the water is not especially deep and there are stones sticking up which, with additional work, should have been possible to use as a kind of ford construction in spite of the rapids.Instead, there may have existed ideas about this place which did not make it suitable for crossing, and perhaps the rapids should be seen as another phenomenon that strengthens the ideas about this place as liminal.The modern explanation for the rapids is that they are created by the changing altitudes of the stream-bed.The otherwise mainly calm and quiet Sagån (the name may be a transformation of the Old Swedish adjective sicever, suver, calm, quiet (suwu strember in the Medieval Law of Uppland) (Ståhl 1985:100)) changes here into small rapids which make the water foam, roar and rush, but not far from here it becomes a calm stream again, what during the Late Roman Iron Age was a narrow arm of the sea.The water probably behaved in the same way during the Late Roman Iron Age, since the water of the higher part of the stream- bed, at an altitude of 10-15 m. a.s.l., falls down into the lower part of the stream- bed at an altitude of 5-10 m. a.s.l.The modern explanation for how the rapids are created would not, however, have been used during the Late Roman lron Age.The explanation during that time instead may have concerned certain powers or other "mythical" (but then most real) phenomena.This place was probably seen as different from many other parts of Sagån, since this was where the stream met the sea; and because of the ideas about the rapids, it was therefore not regarded as a place to cross the stream.
Thus, the place chosen for the stone-setting with the Apollo Grannus vase may have been regarded as a liminal area or a borderland.It was situated on the border between the waters of the familiar and the potentially unknown, but it was apparently also a zone suitable for the remains of a deceased person about to enter another stage of life.The salt water that signified the unfamiliar and foreign by leading to less well-known parts of the physical world, may also have signified other parts of the world than physical ones.Perhaps there was not such a big difference between "physical" and "mythical" parts of the world.Physical places far away may have been seen as almost mythical, and parts of the world that we would regard as mythical may then have been regarded as quite real.A place like the one where Sagån and its rapids met the sea could have been viewed, not as a place suitable for crossing, but as a gate or point of departure for journeys to physical and mythical less known parts of the world as well as for entering another stage of life, which above was compared to both a birth and a journey.The use of the Apollo Grannus vase as a burial urn and the final placing of this vessel by the rapids where Sagån once met the sea, can ultimately be seen as a result of a process of creolization.A blend of local and provincial-Roman ideas concerning Apollo Grannus and this vessel made this specific vessel especially suitable for use in this way at this particular place, but the burial traditions of the area and the local landscape were of course also of importance.
To summarize this point, and also to tie the different threads of this paper together, it can be said that the Apollo Grannus vase was a vessel which, like other vessels of the same period in Scandinavia (Holand 2001), was associated with fertility and rebirth/regeneration.
In this case the symbolism was strengthened, however, through ideas about the vessel based on its Latin inscription mentioning the provincial-Roman god Apollo Grannus, who was associated with fertility (in addition to water).The Apollo Grannus vase was ultimately used as a burial urn, and the fertility and rebirth/regeneration symbolism of the vessel was important for the process of transformation through which the deceased person was "bom" into a new stage of life.Several acts were part of this process of transformation, which in addition to a birth can be compared to a journey, for example the cremation, the placing of the cremated remains in the Apollo Grannus vase, and the placing of the vessel in the grave.The Apollo Grannus vase was originally from a place "far away", and this aspect of the vessel may also have made it suitable to use as a burial urn since the deceased was about to depart for the final journey of life.And what was a better place for the departure of this last journey than the one where different waters met, the water of the well known and the water which would finally lead to the unknown'?
The choice of the Apollo Grannus vase for a grave by the rapids, and not another vessel associated with fertility and rebirth/regeneration, may have depended on the additional ideas about this vessel inspired by the inscription mentioning Apollo Grannus, who was originally associated also with healing springs and their hot and bubbling or moving water.Through a process of creoliza- tion, a connection among Apollo Grannus, the vessel and "energetically moving" water was created, and this connection made the Apollo Grannus vase the most suitable choice for a burial urn at the place where the otherwise calm water of Sagån met the sea in the small, foaming rapids.To clarify further, one can say that another vessel associated with fertility and rebirth/regeneration could have been used as a burial urn at this same place.However, because of the additional associations of fertility and moving water, which were the results of a process of creolization, the choice fell on this particular vessel with the Latin inscription mentioning Apollo Grannus.Another, perhaps complementary, reason for using the Apollo Grannus vase at this particular place near the meeting and mixing of different waters, may have been the former use of it as a vessel for the mixing of drinks.

SOME FINAL WORDS
Around some Late Roman Iron Age graves, cemeteries developed, for example around the so-called grave X at Tuna in Badelunda parish, Västmanland, which contained among other things different kinds of snake-head rings, beads and originally Continental objects (Nylén k, Schönbäck 1994a, 1994b).This does not, however, seem to have been the case with the grave containing the Apollo Grannus vase.The reason for this may be the position of this grave by the rapids in Sagån where the stream met the sea, a position which in the above interpretation of the Apollo Grannus vase is quite significant.The elevation of the land, which caused the point where the stream met the sea to be moved further south, may have implied that the place of the "Apollo Grannus grave" lost its significance, and therefore the reason for continuing a cemetery at this place also disappeared.
A similar reasoning, but the other way around, also explains why this specific vessel was used as a burial urn in a grave at this particular place during this particular time not until then was this a place where different waters met in the rapids.
English revised by Laura Wrang.

ABBREVIATIONS ATA Antiquarian-Topographical
Archives RAÄ + number: Number of registered ancient remain according to the Register of Ancient Monuments at the National Heritage Board SHM + number: Number of registered object at the Museum of National Antiquities Carreni Sn'edieh Archaeology, Vol.II, 2003

Fig. 6 .
Fig. 6.Map showing the area aronnd Fyck- linge with place-names indicating bridges (Broby) and fords (Hanvad, Skars vad) across Sagå n.There are also remai ns ofstone con- strttcti ons regi stered as fords in the area, RAA J78 in Björksta parish, Västmanland, and RAA 236 in Breds parish, Uppland.The map was drawn by the aathor and based on the topographical tnap and the Register ofAncient Monuments.
Carreni S&vedish Archaeology, Voh ll, 2003 THE USE OF THE APOLLO GRANNUS VASE AND ITS PLACEMENT BY THE RAPIDS AS A RESULT OF CREOLIZATION