The Ilam pan
1 It was explained in Chapter 8 that Congavata was the last fort on the Hadrianic frontier, namely Bowness-on-Solway, and that Maia/Mais was the last fort on the Trajanic frontier, namely Kirkbride. What this means of course is that no cup, patera or pan which commemorates either the Trajanic or the Hadrianic frontier can have an inscription including both names Maia and Congavata. It may include one, but not both. And yet the inscription on the Ilam pan includes Mais and Coggabata, the latter evidently intended to be recognisable as Congavata. It has the initial Co and the final ta, but the jumble of letters in-between actually destroys the meaning conveyed by the Notitia name. Usually, however, when one goes back in time the form of a name conveys more meaning, not less. This last point, coupled with the fact that both names Mais and Coggabata appear on the Ilam pan, makes it quite clear that the Ilam pan, no matter how pretty it might be, is not authentic.
2 The inscription on the pan comprises four fort-names, MAIS, COGGABATA, UXELODUNUM and CAMMOGIANNA, together with the rather odd text RIGOREVALIAELIDRACONIS, which by common consent should be subdivided to yield RIGORE VALI AELI DRACONIS. No single source indicates all of the above four fort-names. The spelling of Mais and Uxelodunum is identical to that of the same names as they appear on the Amiens patera. And Coggabata and Cammogianna appear to be modified versions of the Congavata and (C)amboglanna of the Notitia Dignitatum. The original forms of the Notitia names appear to have been Concanata and Cambaglanda, as explained in Chapter 8. Both Concanata and Cambaglanda are straightforward topographical names, again as explained in Chapter 8. The form Coggabata on the Ilam pan does not appear to have developed from the original name – a direct change from the n of Concanata to the b of Coggabata seems unlikely, though a first change to the v of Congavata and then a further change to the b of Coggabata is entirely possible. Likewise a direct change from the nc of Concanata to the gg of Coggabata seems highly unlikely, whereas a change from the ng of Congavata to the gg of Coggabata is at least possible, even if somewhat artificial and contrived – a change to nn would seem much more natural since Connabata is easier to pronounce than Coggabata. The same appears to be true of Cammogianna. A direct change from the original Cam bagl anda to the Cam mogi anna on the Ilam pan seems most unlikely, whereas the change from (C)amboglanna to Cammogianna is at least within the bounds of what is credible. The mb of the Notitia form has been simplified to mm and the l has been replaced by i. But note that the engraver of the Ilam pan has carefully preserved the beginning and end of each of Congavata and (C)amboglanna – it is in the interior of each of those names that he has given free rein to his creative urges.
3 The engraver appears to have taken Mais and Uxelodunum (both copied correctly) from one source, presumably the Amiens patera (which does not include Congavata) and to have taken Congavata and (C)amboglanna (both modified) from a different source, presumably the Notitia Dignitatum (which does not include Mais and Uxelodunum). He has then apparently arranged these two pairs of names taken from different sources in a single list on the assumption that Mais (taken from one source) was the last fort on Hadrian’s Wall and that Congavata (taken from the other source) was the next fort to the east. Logically, given that assumption, the Ilam pan cannot then be used as proof that Mais was the last fort on the wall and that Congavata was the next fort to the east. It follows that the Ilam pan gives no useful information not already available from other sources. Worse still, it is actually misleading.
4 Having regard to all that is said above, it seems quite clear that Congavata and (C)amboglanna did not develop naturally into Coggabata and Cammogianna, and that these last two names are not the result of miscopying, but that Congavata and (C)amboglanna have been deliberately modified to give the Coggabata and Cammogianna on the Ilam pan. This is no doubt a sign that one should not take RIGOREVALIAELIDRACONIS too seriously, since this text, too, is probably the result of modifying some other text, a text having nothing to do with Hadrian’s Wall or with a Mr. Draco or Draconis, and evidently not having a third letter l which would have enabled the engraver to spell VALLI correctly. In other words, RIGOREVALIAELIDRACONIS is an anagram. Now, there are no doubt several possible re-arrangements of the letters of that text which would yield a meaningful original text, but the re-arrangement which appears to be correct is LEONARD RIVER ISO GALICIA. There is indeed a river Iso in Galicia, in A Coruña province. This illustrates a remarkable achievement on the part of the engraver, for just as AELI on the pan can go with RIGORE VALI to form an expression representing a linear geographic feature – Hadrian’s Wall – or with DRACONIS to form a personal name, so in the original text the word RIVER can go with ISO GALICIA to form an expression with a linear geographic significance – a river in Galicia – or with LEONARD to form a personal name. That is no mean achievement, and one can imagine the engraver having many sleepless nights working on his anagram! But we are not yet finished, for one of the foremost scholars in the field of Roman place-name studies was the late A.L.F. Rivet, referred to at numerous points in the present study. He spent a great part of his life working on the etymology of place-names and it so happens that the modern English word ‘rivet’ is derived from Old French ‘river’. It does therefore seem quite clear that LEONARD RIVER is merely a thinly veiled reference to Leo Rivet, the name by which A.L.F. Rivet was apparently known to his friends and colleagues.
5 In light of the above it may be that the inscription on the Ilam pan is just some kind of university prank, it being noted that the village of Ilam, where the pan was found, is only some 16 miles from the university where Rivet worked for many years.
6 There are various points to note regarding the inscription on the pan, including the rather inelegant use of an incised guideline at the top of the letters, the fact that some letters encroach upon the decorated frieze of the pan, the uneven spacing and height of the letters and perhaps above all the rather sloppy execution of the NIS of DRACONIS. The N is not even vertical and the S is quite different in form from that of MAIS. All of these points can be seen as clues that the engraver was an amateur and as hints that we should not take the engraving too seriously.
7 What is not clear is whether Rivet himself played any part in the prank or whether others used his name without his knowledge, perhaps even after his death. But what seems absolutely clear is that the Ilam pan, contrary to the view of Jackson (Jackson 2012) and others, is most certainly not a souvenir from Hadrian’s Wall, and its inscription is of course of no relevance whatsoever to Roman place-name studies.