Discussion The Channel of Corbulo itself , while directly flowing into the Rhine next to Matilo , may not have been a Roman world border , but it can be argued that it was a liminal border in-between cosmological worlds . The mask was taken out of use by depositing it in a place where it could not be retrieved or seen anymore ( cf . Cousins 2014 , 61 ; cf . Kamash 2009 , 232 ). On top of that the Corbulo channel was a converted natural waterway , meaning that the fuzziness whether it was seen as natural or artificial was connected to the deposition as well ( cf . Cousins 2014 ; 6061 ). Therefore , an explanation for a ritual deposition of the Matilo mask and the associated finds in terms of liminality seems plausible . The channel , being a liminal space , would have functioned as an extraordinary place between ( cosmological ) worlds , where objects could be offered or ‘ move through the in-between and leave it behind with a difference ’ ( cf . Derks 1998 , 132 ; cf . Fokkens and Fontijn 2013 , 564 ; cf . Kamash 2009 , 230-232 ; cf . Nicolay 2007 , 124 ; cf . Thomassen 2015 , 40 ).
The mask was found at the bottom of the channel ( fig . 2 ). A couple of meters away from the mask , heavily chopped articulated horse bones were found that belonged to two horses ( Lauwerier and Robeerst 1998 , 22 ). In addition , the skull of a horse was found . Research on this skull revealed that the horse was decapitated and that the chopping of the other bones was not done for functional butchering reasons ( Lauwerier and Robeerst 1998 , 13 ; 22 ). The bones of a leg were found together , while normally these bones would disperse after butchery ( Lauwerier and Robeerst 1998 , 21 ). Thus , Lauwerier and Robeerst ( 1998 , 23 ) propose that a ritual following the pars pro toto principle had taken place : the parts of the horses representing entire animals . Especially when it comes to the horse skull , the ritual resembles the find of the mask : both horseman and horse are represented by the facial area . The mask was broken off its helmet and the horseman had armour which was not found either : both are arguments for a pars pro toto explanation of the ritual . However , the incomplete state of the mask might as well have been a reason for its discard . Still , the mask and horse bones differ in another respect : the individual parts of the horse skeleton were destroyed on purpose , whereas the mask was not ( cf . Lauwerier and Robeerst 1998 , 23 ; fig . 2 ). The act of damaging an object has long been perceived to signal the end of its use or meaning ( i . e . ritual killing ). Nevertheless , recent work on broken objects shows that is possible that fragmented finds served new purposes and embodied the whole ( Croxford 2003 , 82-83 ; 93 ). The same might apply to these fragmented finds which had now become part of a ritual , possibly serving a new purpose .
Considering the horse bones were classified as ritual and that many finds from the mask layer like the weapons , horse gear , and coins have often been found in ritual deposits it can be argued that it is possible that ( at least a part ) of the find assemblage was deposited in the channel in a ritual ( cf . Martens 2004 , 142-144 ; cf . Nicolay 2007 , 85 ; cf . Roymans 1990 , 77-80 ).
Furthermore , the rarity of mask finds and the mask ’ s value also point to a ritual deposition ( cf . Hazenberg 1997 , 38-39 ). Since , the context date matches the mask date the Matilo mask was removed from the world in a period in which it was still in common use , indicating the mask ( with a repaired hinge ) could have still served its original purpose . And , because the find context of the mask is likely to be unimpaired based on the articulated horse bones , suggesting the finds did not end up in the channel by chance , it is possible to interpret these finds as having been ritually deposited in a liminal place ( cf . Lauwerier and Robeerst 1998 , 21 ).
Conclusion The coincidence of the dating range of the Matilo mask and associated finds , the rarity of the mask , the ritual deposition of the horse bones , the other potentially ritual finds , and the undisturbed context of the Matilo mask all make it plausible that the mask was deposited in the Corbulo channel as a ritual action .
The finds from the mask layer and the mask itself appear to have been deliberately deposited . As liminality can be attributed to the Corbulo channel this may have been sufficient reason to render the channel suitable as an extraordinary place for depositions . Although this explanation in terms of liminality and the current model , in which soldiers offered equipment in rites of passage to Gods in return for their safety , are not mutually exclusive , it can be seen as an alternative or complementary interpretation .
In future archaeological research on Roman wet contexts it is important to be more cautious when interpreting the find categories , which were discussed in this article , as normal waste . And , it seems liminality might be needed as an additional dimension to explain why depositions of Roman helmets took place in wet contexts along the Limes .