One of the most pervasive problems in microscopy is that of making sure that what one is observing represents reality. The complex specimen preparation steps which have to be followed to obtainimages of biological samples all produce artifacts, at least in the sense that they modify the sample. The investigator has to somehow show that these changes in the sample reveal true microscopic features, rather than structures which do not exist in the living. This, of course, is why polarized light microscopy, phase and interference microscopy and their derived approaches, all of which reveal detail in live untreated specimen, are so important, in spite of the limited resolution which can be achieved.
When electron microscopy was in its infancy there were, and justifiably so, even more severe fears about the reality of what could be seen. Even visionary people like Gabor had their doubts and the first steps in biological microscopy were taken with the realization that if the sample was burned to a crisp, perhaps examination of its cinders would still give valuable information.