Problems of Electron Stereoscopy of Biological Tissue

A detailed study of stereoscopy of tissue sections with the transmission electron microscope has been carried out using different tilt angles and staining techniques, aided with models. Pictures with tilt angle difference of 20 or more can be fused to give pairs of micrographs showing stereodepth. Cytoplasmic filaments, glycogen and ribosomal granules give good stereodepth. Oblique membranes, especially when stained with lead or phosphotungstic acid, give good depth in stereo-pairs but membranes running more vertically in the section do not. A theory, based on the ' transparency factor' that is a feature of transmission electron microscopy, is given to account for this discrepancy. A learning process is involved in achieving depth perception when viewing stereo-pairs of electron micrographs of tissue sections. Complex structures often cannot be perceived in depth even after prolonged scrutiny through the stereoscope. Where structures overlap in the thickness of the section, they can be 'separated out' with stereoscopy and in this way clarified, for depending on which micrograph of a stereo-pair is presented to which eye, a structure can be made to appear either in the front or the back of a section. In general, problems of depth perception in stereo-pairs of micrographs obtained with transmission electron microscopy arise because with our normal everyday vision using reflected light, we look at things, whereas with the electron microscope using a transmission electron beam we look through things.