After scraping the epithelium from the rabbit's cornea, pseudoeosinophilic leucocytes appeared in the limbic area within one hour and began to advance, presumably by means of pseudopodia, into the anterior corneal stroma. The cytoplasmic granules of these cells were intensely stained by eosin and by the Undritz peroxidase method during the first hour, but were not stained by the PAS method until the eighteenth hour. Electron microscopy failed to show crystalloids in the granules. The cellular changes were followed until the fourth day, by which time most of the cells had disappeared by fragmentation, karyorrhexis and lysis.