SYMBIOSIS BETWEEN TERMITE AND FLAGELLATES. 317 



end of the eighth day of starvation. In the oxygenation experi- 

 ments, Trichomonas disappeared entirely within 24 hours. It 

 was obvious, then, that a combination of these two methods for 

 removing the protozoa would yield interesting results. Accord- 

 ingly, ten experiments, using approximately 400 termites, were 

 carried out as follows: The termites were oxygenated, as in 

 previous experiments, for 24 hours. This removed all protozoa 

 belonging to the genus Trichomonas. About 100 of these ter- 

 mites were starved for six days, and 100 for eight day-. In 

 this manner, the lir>t loo, or those individuals which were 

 starved six days, were freed of Trichonympha. and the second 

 IOO, or those individuals which were starved eight day-, \\ere 

 freed nf Trillion ymp ha and Leidyopsis. Thu>, in those indi- 

 viduals that were starved eight days, Streblomastix only re- 

 mained, while in those individuals that were starved six da\ >, 

 and Streblomastix remained. 



d. ll'oixl-'oiling after Intervals of Starvation and Oxygenation. 



By >tai -\ation and by oxygenation and by a combination of 

 -larxation and oxygenation we have seen how it is possible to 

 >hift tin- protozoa about almost any way we wi-h. For instance. 

 \\e i .in t.ikc out Trickonympka by starving MX days and lea\ e 

 /.ciilyof'sis, Trichomonas and Streblomastix uninjured; by starv- 

 ing eight days we can remove Trichonympha and Leidyopsis, 

 lr.i\in^ Trichomonas and Streblomastix, and then by oxygenating 

 .?4 honi> we can remo\'e Trichomonas, leaving only Streblomastix', 

 or \\r in.iN 1 1\\ senate tirst and remove Trichomonas which \\ill 

 lea\e Trithonynjpha. Leidyopsis and Streblomastix, and then if 

 \\e starve the-e individuals for six days we have Leidyopsis and 

 !>lmnastix remaining; and by oxygenating for ~2 hours all 

 I'H'io/n.i are removed. By this crisscross procedure we may 

 obtain termite:- \\ith live 1 protozoal combinations and protozoa- 

 less termites a> follows: (i) Leidyopsis, Trichomonas, Streblo- 

 nnistix; (2) Trichomonas, Streblomastix; (3) Streblomastix; (4) 

 Trichonympha, Leidyopsis, Streblomastix; (5) Leidyopsis, Streblo- 

 mastix; and (6) no protozoa. Thus a wonderful opportunity is 

 afforded for studying the relation of each ot the four genera of 



1 Two more combinations have been obtained recently. See addenda for a tab- 

 ulation of all combination-. 



