I am particularly interested in western European single handed swords that developed between the late 10th to mid 14th centuries and have spent quite some time researching them. In doing this research I have found a major discrepancy between the prevalence of wheel pommels when depicted in art works and found in surviving specimens. Artworks from the 11th century onward depict them very commonly and would lead one to think that by the 12th century it was the most common form. But from what I've found so far while looking into the archeological specimens this is not the case at all.

The earliest wheel pommels I'm aware date to the beggining of the period in question (late 10th century) however I've found that they don't feature commonly on surviving specimens until the late 13th century. Until this point the various "nut" forms such as the tea cozy, Brazil nut and cocked hat are the most common along with many bizzare and unique forms that don't conform to any particular style.

Has anyone noticed this or can provide some kind of explanation why I might find this contradiction?

Well, this may be reductive, and a bit silly, but those manuscripts were really tiny; I think it'd be pretty hard to draw a cocked hat, and would look messy and not make a lot of sense on the page.

Sam,

I have made the same observation. I think, in part, it might be because central Europe, particularly the Empire, had a well-developed sword manufacturing industry. For whatever reasons, Brazil nut pommels and other variants remained in vogue significantly longer than in other parts of Europe. We have manuscript evidence, if not artifact evidence, for Brazil nut family pommels even as late as circa 1350 AD. So the German influence could be a factor.

Additionally, depending upon what books you own and web resources you browse, you might have a distorted perception of the number of surviving disc and wheel pommel swords. For example, have you seen all of the antique swords in this thread?
http://myArmoury.com/talk/viewtopic.php?t=33337&highlight=

I suspect these explanations cannot entirely account for the phenomenon you have described, but might provide at least a partial explanation.

Two big factors, preservation bias and collector bias. The frequency you find stuff in the ground or in collections doesn't necessarily reflect their relative popularity in their working life.

I would suggest an additional aspect to this issue. Today we see these pieces as precious artifacts to be treasured and emulated. (here probably even more than the general public :) ) in the time of the swords use they where much more tools and weapons to be used and modified as needed. Wheel pommels in style longer than the brazil nuts for example could be reused and refitted to new blades. The result would be fewer dateable examples to the earlier period in the population of items that are not ground finds.

Best
Craig

Interestingly enough there are many surviving examples of wheel or disk pommels that are slightly flattened. That could be an artifact of some aspect of their manufacture but it can just as easily result from grinding down the peen when no peen block is present. That said I did take down a Windlass Type XIV once and it was surprising how little metal had to be removed, hardly altered the shape of the pommel at all.

Anyway has somebody done a rigorous count of nut vs wheel pommels on surviving swords to establish whether or not the OP's observation might even be a product of sample bias? There are a large number of pilzform pommels out there but I'm also aware of a large number of early disk pommeled swords with find sites going all the way to Russia.