Assuming that this was a trail made by an aircraft flying at 30,000 feet or above (ie contrail level), then my estimate was that the wavelength (ie "crest peak" to "crest peak") of its sinusoidal course was 2-3 miles, and the amplitude (ie "crest" to "trough") about a mile. However it might of course have been considerably higher, particularly if military, in which case those estimates could double. From where I was observing (Chertsey) I would estimate that the sinusoidal track had been carried out for at least 30 miles. I can't believe that this could be a result of normal autopilot deviations, albeit within airway constraints - the pilots would have become aware of it rather early on, methinks.
What a shame that we have no posts from anyone who saw the aircraft making it, and none of us had cameras handy either!
Incidentally the "blobs" mentioned by several posters seem to me to be a frequent effect during the breakup of normal contrails as well. Although there were some apparent on this one, I was struck by how tight most of the wavy trail remained until it finally evaporated.