All the species examined belonging to the Mastotermitidae, Kalotermitidae, and Rhinotermitidae harbor a teeming menagerie of intestinal protozoa. Most species of the Termitidae examined had no protozoa, but some of these give over a portion of their nest to the growth and cultivation of a fungus which they eat along with soil, cellulose, and hemicellulose upon which the fungus acts. When they eat wood it is usually much more decayed than that which the protozoa-harboring families eat. If they can digest cellulose it is probably because of certain fungi or spirochaetes which they harbor in their intestines. A number of individuals of Reticulitermes were freed of their intestinal protozoa by incubation at 36[degree]C. for 24 hrs. Such termites usually die within 3 weeks. The actual length of time required for death, however, depends upon the food eaten; the more decayed the wood, the longer they lived. The fact that the incubated termites lived much longer, if not indefinitely, on a predigested diet of humus or fungus-digested cullulose than on their normal diet of partially decayed wood, indicates that the incubation itself was not responsible for their death. When the protozoa were restored the incubated termites were able to eat their normal diet of wood and live on indefinitely. The termite and its protozoa, then, are true partners or symbiotes; the termite chews and swallows wood, which the myriads of protozoa in its intestine digest for themselves and for the termite that shelters them. By the use of starvation and oxygenation methods, selective elimination of species of the intestinal protozoa may be accomplished. Enough of this work has already been done to demonstrate that certain species (Trichonympha, Leidyopsis) are more valuable to the host than others (Trichomonas), while some (Streblomastix) are of no value.