Reflected light microscopy of biological material has been a very difficult task and many different but hardly successful attempts have been made to get usable images. The main reasons for this state are: weak reflections from the biologically important structures in the object as against strong reflection at its surface; reflections at optical surfaces in the microscope which cause a deterioration of contrast; and the mixing together of reflections from many levels in the object so that the signal coming from the focussed-on level is lost in noise and d.c. components. We tried, and successfully, to reverse this situation and to make it possible for the signal to overwhelm spurious and scattered light using double, or, as we call it, tandem scanning. The object is illuminated only in small patches lying in one plane and moving across this plane and only the light reflected from these illuminated patches is allowed to pass into the image plane and participate in image formation. The image consists of points which travel over the image plane and which are geometrical images of the illuminated patches in the focussed-on object plane. To ensure that only the light belonging to these geometrical images be allowed to enter the image, the image plane is covered with an opaque diaphragm having holes in locations exactly corresponding to the locations of the geometrical images of the illuminated object points; this diaphragm travels in concordance with the first scanning and so the complete image of the focussed-on object plane is formed successively. Both scans are performed by a single device, a Nip-kow disc carrying in its annular periphery several ten thousands of holes arranged in Archimedean spirals. The disc is 100 mm in diameter and rotates about 100 rpm driven by an electric motor. On one side the disc is illuminated in a circle 18 mm in diameter, and the light transmitted through several hundred holes and reflected in a mirror system passes a microscope objective which forms the images of the disc holes in the object plane. The light reflected here passes through the same objective and mirror system (one mirror being a 'limitlessly thin* beam splitter) to pass conjugate aperture holes on the observation side of the Nipkow disc. As the disc lies at the intermediate image plane of the objective lens on both its illuminating and observation sides, only light emanating from reflection or fluorescence in the plane of focus can contribute to the image. High contrast images of very thin focussed-on layers are thus formed. The practical arrangements are such that very large specimens can be examined: the specimens for this microscope need not themselves be microscopic.