It's great knowing Tim - he's even more obsessional than I am
To his credit he didn't let this go until he resolved the dilemma. In good deductive style, Tim was able to establish that these 'smudges' on the edges of thumbnails occur when there is an image that is cropped in such a way that the subject is against one side of the image, has varying colours/contrasts and the other side of the image is 'open'. What then seems to occur is that a coloured bit of the image against the dark of one edge is mirrored by a kind of 'negative' or 'reciprocal' colour on the opposite open edge of the thumbnail - e.g. light on one side, dark on the other.
To illustrate, I trawled the whole database for a good example, and just happened to find one of my own photos
View Large View Medium
Photo © Paul Markman
Here you can see that the strips of colour intersecting the right hand edge of the image are replicated in a kind of negative/reciprocal colour on the left hand edge.
All fascinating stuff, and potentially completely irrelevant
. As has already been said, this feature of the photo is not replicated in the larger images. I have tried to reassure Tim that he doesn't need to worry about this any longer and I understand he shall be discharged from therapy shortly
. Seriously - I admire such an enquiring mind. Well done Tim for solving this challenge.