286 L. R. CLKYI I AND. 



The Successive Stages in Food Ingestion. A flagellate just 

 previous to taking in food (Fig. i) usually has few wood particles 

 inside the body; whereas at other times about one third (the 

 posterior third) of the body is always filled with the wood par- 

 ticles. First the longitudinal myonemes contract, thus causing 

 the posterior end to become more rounded and slightly flattened 

 centrally (Fig. 2). It is quite possible that the transverse 

 myonemes expand at the time, though it has been impossible to 

 actually observe this. In the next stage (Fig. 3) the body is 

 more contracted and hence shorter; this contraction, owing to 

 the fact that the inner or central portion was pulled forward while 

 the outer and more rigid ectoplasmic portion did not invaginate, 

 results in the formation of a cup-shaped cavity (Fig. 5) lined 

 with the wood particles which adhered to the outer and un- 

 differentiated region prior to invagination. Now the cytoplasm 

 of the most posterior portion of the undifferentiated region soon 

 begins to flow backward, thus narrowing the posterior end of 

 the cavity (Fig. 6). This continues until finally the cavity is 

 completely closed, the wood particles remaining inside (Fig. /) 



Other Trichonympha sometime get caught in the cavity just 

 as it begins to close (Fig. 9), and may be seen swimming in this 

 position for ten minutes or more, for caught individuals free 

 themselves with great difficulty. This, of course, is not a com- 

 mon occurrence, though five such observations were made within 

 four hours. In three termites, out of more than 500 that were 

 examined, perhaps half the Trichonympha present had other 

 Trichonympha partially ingested and securely fastened inside 

 their bodies. The cause of this peculiar phenomenon is entirely 

 unknown. Living individuals of Trichomonas terniopsidis, a much 

 smaller flagellate always associated with Trichonympha, are some- 

 times ingested along with wood particles, though not in sufficient 

 quantity perhaps to be of much food value. Leidyopsis sphccrica, 

 another protozoon closely related to and associated with Tricho- 

 nympha in Termopsis angusticoUis, takes in wood particles simi- 

 larly to the method described here. It has also been observed 

 killed and half ingested by Trichonympha. This, however, is a 

 rare observation. 



