L. R. CLKYKI.AXD. 



the others, and the experiments were examined one by one until 

 the minimum was found to be within certain limits. Then the 

 experiments were set up again and were examined at various 

 intervals within the known limit. Sometimes it was necessary to 

 repeat the process many times, before a fairly accurate minimum 

 was finally determined. The minimum thus determined was then 

 tested out three times. 



Termites were usually examined immediately after having been 

 removed from oxygen and non-motility was the criterion for de- 

 termining whether or not their protozoa had been killed. It was 

 found that if a few protozoa were motile at the end of the oxygen- 

 ation period they did not die later. Some hosts which had been 

 freed of motile flagellates were examined at intervals up to ten 

 days, and no protozoa ever appeared in any of them. It is of 

 interest to note that the protozoa disappeared from the intestine 

 very soon usually within two to four hours after they had been 

 killed. They were probably digested by the termites. 



At one atmosphere the protozoa of Cryptotermes and Reticuli- 

 termes were not all killed in all hosts in ten days; however, they 

 were all killed in a few hosts even in three days. They were all 

 killed in a great majority of hosts in ten days, but in a small 

 number perhaps about 5 per cent. -some protozoa were alive at 

 the end of ten days. 



The protozoa of Leucotermes were killed very much more 

 quickly than those of the other termites until a pressure of 2.5 

 atmospheres was reached. These differences in oxygen toxicity 

 are not correlated with size of termite hosts, for Termopsis is 

 approximately twenty times as large as Leucotermes and ten times 

 as large as Cryptotermes, but Reticulitermes and Leucotermes are 

 about the same size. Difference in habit may be a factor, but 

 Reticulitermes and Leucotermes are very similar in habit as well as 

 in structure. The protozoa of these four termites, although all 

 flagellates, are nevertheless quite distinct morphologically, many 

 of them belonging to separate families. Hence, it is possible that 

 the differences in oxygen toxicity may be found to be in the 

 protozoa themselves. 



In Termopsis, as was true in previous experiments (Cleveland, 

 '256), Trichomonas was killed first and Trichonympha last until a 

 pressure of three atmospheres was reached. Then a peculiarly 



