It is interesting what individuals do for their microscopic images for publications. Looking over all the TEMs i can find for SP-D, SP-A, conglutinin, MBP, etc I find that some methods for presenting images are nice, they probably are NOT what I would have done.  In 1993 when this manuscript went to press I don’t believe that very many individuals recognized the potential for photoshop to create “pretty” images.

Changing the contrast by dodging with fingers and spoons and holey circles was about the extent that an image could be manipulated in the wet dark room.  Occasionally we huffed and  puffed on a print to warm parts hoping for more development on one side or another but other than that, there wasn’t much one could do with a print to beautify it.

In this manuscript (Eur. J. Biochem. 215, 793-799 (1993)) the authors have performed (in my opinion) the interesting, unnecessary, somewhat deceptive task of putting three SP-D molecules (the dodecamers) into a straight line portion B of their figure. It didn’t take much “looking” to see that the same SP-D dodecamer is present in a field of their part A of that figure, and again with two partner dodecamers in part B of the same figure. While this doesn’t necessarily change their data, one was rotated slightly from the other, and just a tad different in magnification. The better approach might have been to put a white division line between the first, the second and the third SP-D dodecamer in panel B, rather than to fuzz the background together…. that was just on the verge of being deceitful, since it gives the impression that lots of routine dodecamers appear close together in their prep. Probably not the case. That said. Below is their figure part A with the single dodecamer, and part B where they have cut pasted and blended the shadow background. You judge.

It might be noted that journals right about this time built (or were supposed to have built) algorithms to detect manipulation of the pixels in publication images.
And just in case you are skeptical that they are the same image, see the radial pixel pattern upon rotating half transparent images one from part A and one from part B just a few degrees (which doesn’t happen on non duplicate images (again, in my experience)).