SYNOPSIS.The conclusion drawn in 1921 that the large nuclei in the cytoplasmic cortex ofGlugeacysts are not vegetative nuclei of the microsporidan but nuclei of the hypertrophied host cell was based on the discovery of early developmental stages in the mesenchyme of stickleback larvae experimentally fedGlugeaspores. This observation had been made on serial sections from experiments done in 1912. The intracellular development of the microsporidan could be followed up in this material only thru the 1st stages of schizogony. Renewed infection experiments, done still in 1921 on a much broader basis, have fully confirmed the previous findings, as briefly stated in 1922. On this material, the intracellular development ofG. anomalahas been followed up in recent years from uninucleate host cells 7 μ in diameter, interpreted as wandering cells in the mesenchyme, until they became macroscopic multinucleate cysts, in which schizogony and sporogony of the microsporidan produced innumerable vegetative stages and spores ofGlugea.The details of the developmental processes are described in the present paper.The multinucleate host cell and the intracellular parasites together form one of the symbiotic complexes for which the term “xenom” or “xenoma” has been used by me since 1949. By a sequence of amitotic nuclear divisions, the uninucleate host cell in theGlugeaxenomas ofGasterosteusbecomes plurinucleate in contrast to the usual structure of other xenomas of fish.Already in 1921, I thought that the host cell in theGlugeaxenomas may have phagocytic properties. The observation of accumulation of granules from pigment cells in some of theGlugeaxenomas has now verified this sup