Automation of the acquisition and interpretation of data in microscopy has been a focus of biomedical research for almost a decade. In spite of many serious mechanical perception of microscopic fields with a reliability that would inspire routine application still eludes us. Many facets of the problem appear to be well within the grasp of presentday technology. Thus, available histochemical techniques make it possible to prepare biological materials so that morphological integrity is preserved, key constituents are stained stoichiometrically, and the specimens are favorably dispersed for effective imaging one by one. Scanning microscopes now have the requisite sensitivity, resolution, and stability to sample such objects and make photometric measurements over a wide range of magnifications and wavelengths within the visible and near-visible spectrum. Furthermore, modern large capacity, high speed data facilities at last provide the ability to manipulate the hitherto unmanageable quantities of optical information contained within all but the simplest images. With the basic materials for achieving automation via mensuration finally at hand, attention has been turned toward generating and evaluating methods for extracting meaning from quantitative optical information. Definitive concepts for image structuring and image characterization have yet to be realized, to be given satisfactory operational definitions, and to be assembled within a machine-oriented perceptual framework. 6 Criteria for effective and efficient discrimination and interpretation of images must be evolved. It would be a serious mistake to presuppose that mechanical perception must mimic the human’s perceptual apparatus in organizing images as complexes of picture eIements. Likewise it would be serious to ignore traditional descriptive morphology and established taxonomies. In steering a middle course, the exploration of many complementary approaches and the introduction of numeric methods which fully utilize measurements of light intensity of optical density seemed to us to hold the most promise for augmenting and explicating the existing, largely verbal tradition of microscopic morphology. To realize these objectives, we have designed a system with three basic components: 1) a sensor which rigorously and effectively scans microscopic fields and converts the optical information into digital form; 2) human analysts who contribute heuristics, devise image processing methods and en-