"The "maes" element always has been rather puzzling, it doesn't seem to be Norse. I can't recollect what Hugh Marwick made of it. He strongly believed in a P-Celtic substratum to Orkney names, and yet he had difficulty producing immediately convincing examples. Interestingly, Jakob Jakobsen, the Faeroese scholar who studied the place-names of Shetland, equally well convinved himself - if no-one else - that the Celtic substratum was a Q-Celtic one. For many years the standard opinion on the pre-Norse language of the North was that promulgated by Prof Kenneth Jackson in his essay "The Pictish Language" in the seminal F T Wainwright volume "The Problem of the Picts" 1955 - that whereas the Picts of Fife, Angus, Perthshire spoke a P-Celtic language (with hints that this may have been slightly closer to Gaulish than to welsh), those north of the Highlands spoke a non-Indo-European language. You can't of course look for traces of a language you don't know. Jacksons view still seems to me to make good sense but the non-IE Pictish has now been challenged by K Forsyth who thinks that P-Celtic was spoken throughout the Pictish region. I can't help thinking though that if it had been so very general, the place-name traces would be obvious - as they are in the southern Pictish area, with "lin" instead of "loch", "lan" for an enclosure around a church, and abundant instances of "aber". If Welsh place-name elements (like the river names "Pant" and "Cam") can survive in East Anglia - a region heavily Anglo-Saxonised - surely some would be apparent in Orkney?

Anyway, obscure elements like "maes" could be non-IE, but no-one can prove they are. It isn't unique to Maeshowe. There is a chambered tomb, as it happens of Maeshowe type, on Start Point in Sanday - its name has been corrupted in recent times to "Mount Misery" but the older form is "Maesry". I don't think anyone can make any more of it than this."