Problems involving the so-called "intracellular symbionts"2 have long troubled some cytologists and bacteriologically inclined zoologists. The solution of many of those problems, however, yet evade even the most careful investigators. Several workers in this field (Buchner, Glaser, Mlercier, Schwartz) hold to the view that the physiology and systematic position of the intracellular bacteroids can be accurately determnined only by the cultivation of these organisms on artificial media. Toward this end, many workers (see Schw^artz, 1935) have struggled with sterilization and dissection techniques and experimented with various culture media in what has usually been futile or unsatisfactory attempts to grow the bacteroids in unnatural conditions. The extracellular "symbionts" in the gut of insects are generally not strict in their growth requirements (Schwvartz, 1935; Steinhauis, 1941), and some of the forms that have an intracellular stage in the gut wall and an extracellular stage in the gut lumen seem to grow readily on artificial media from their extracellular stage: e.g., those of species of Sitodrepa, Ernobius, Rhagium, Xestobium (Heitz, 1927; Muller, 1934), Rhodnius (Wigglesworth, 1936), and others (see Schwartz, 1935, p. 398). The more closely adapted intracellular organisms, however, grow with difficulty if at all, and none of the reports of successful cultures of these are above serious criticism on matters of sterilization and manipulation. The intracellular bacteroids of the cockroaches, w'ith which this paper deals specifically, have been so thoroughly described (Fraenkel, 1921; Gier, 1936; Bode, 1936; Hoover, 1945) that it suffices here to say that these bacteroids are nonmotile, nonsporeforming, faintly gram-positive, straight or slightly curved rods, 0.8-1 ,u by 1.5-6.5 ,, and may stain barred, somewhat like diphtheroids. They occur around the ovarian eggs and in specialized cells, the mycetocytes of the abdominal fat bodies. 1 The work reported here was begun at Indiana University under the direction of Dr. Fernandus Payne, as part, of a doctorate problem, and wAas continued at Harvard University Biological Laboratories on a National Research Council Fellowship and a Harvard University Research Fellowship under the general supervision of Dr. L. iR. Cleveland and Dr. A. B. Dawson. Aid and counsel which made this work possible and profitable are gratefully acknowledged. 2 The term "symbiont" definitely connotes a helpful association between two types of organisms. Such a relationship has never been demonstrated for any of the true "intracellular symbionts." MIercier (1907) designated the bodies of the roach mycetocytes as "bacteroidi" and this lead was followed by Hertig (1921), Gier (1936), and Hoover (1945). It seems better thus openly to confess our ignorance as to the nature of these bodies by continuing to designate them merely as "bacterialike" than to imply a relationship which probably does not, exist, even though these "bacteroidi" may sometime be definitely placed taxonomically with the bacte...