31 8 L. R. CLEVELAND. 



protozoa to its host and to its neighbors or fellow protozoa, for 

 none of these six groups of termites was injured in the least by 

 the methods employed in removing the protozoa. Each group 

 would feed on wood just as it did before the protozoal alterations 

 were made. 



Each of these six groups was now fed wood and kept to itself 

 in the same environment of temperature, moisture, and light. 

 Each group contained about fifty individuals. Controls, or 

 termites that had not been treated in any way, were also kept 

 with these five groups. The results of feeding wood to each 

 group may be briefly stated as follows: 



(1) Termites "with Leidyopsis, Trichomonas and Streblomastix. 

 In normal termites in nature, Trichonympha for some reason is 

 perhaps 1000 times as numerous as Leidyopsis, and this ratio is 

 fairly constant, although we do not know what makes it so. 

 Trichonympha is the dominant genus in size at least and probably 

 in number too Trichomonas and Streblomastix may sometimes 

 be as numerous as Trichonympha, but they are much smaller. 

 A most interesting thing happens to Leidyopsis when its dominant 

 neighbor, Trichonympha, is killed; it multiplies rapidly, soon 

 increases very greatly, indeed, in number, and in 20-30 days 

 has filled up the space made vacant in the termite's intestine by 

 the removal of Trichonympha. This condition remains perma- 

 nently and the group of termites is able to live indefinitely on a 

 wood diet. Leidyopsis, then, is not only able to take the place 

 of the dominant Trichonympha in number but can also take its 

 place as the chief symbiont. As we shall see when we come to 

 study the group of termites with Trichomonas ai\d Streblomastix, 

 or the group with Streblomastix, Trichonympha, in nature, is by 

 far the most important symbiont. We know this, even though 

 we were not able to get a pure culture of this genus, because 

 Leidyopsis is present in too small a number to be of much im- 

 portance and Trichomonas, as group (3) shows, is not of very 

 great moment as a symbiont. So, in nature, Trichonympha is 

 of most value to termites, because for some reason it is dominant 

 over Leidyopsis; although, under experimental conditions, Leidy- 

 opsis can become of as great value to its host as Trichonympha 

 is in nature. 



(2) Termites -with Trichomonas and Streblomastix. When 



